version 5.4.0

CMD packets arrive as unauthenticated UDP datagrams, yet every hbc client
executed the shell command they carry without any opt-in. Add an
allow_remote_command config key, default false: when off, the command is
logged and refused with "Refused: allow_remote_command is false" (visible in
the server event log under the command service), and subprocess is never
reached. When on, the client warns at startup that it will execute CMD
packets.

Applied to all four clients that handle CMD — hbc, hbc_windows.py,
hbc_mini.py, and the C hbc_mini — since gating only one leaves the others
wide open. The C client reads the same key from ~/.hbc.json and needs a
rebuild to pick it up. UPD (self-update) is unchanged.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

.gitea/workflows/release.yml

@@ -10,36 +10,48 @@ jobs:
     steps:
       - name: Checkout code
         uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
         
-#      - name: Set up Python
-#        uses: actions/setup-python@v5
-#        with:
-#          python-version: '3.11'
       - name: Set up Python
-        # Use a generic run step for FreeBSD if actions/setup-python
-        # fails in restricted environments.
         run: |
           python3 --version
           python3 -m ensurepip --upgrade
           
       - name: Install build tools
         run: |
-          python3 -m pip install --upgrade pip
+          python3 -m venv .venv
-          python3 -m pip install build twine
+          .venv/bin/pip install --upgrade pip
+          .venv/bin/pip install build twine
           
       - name: Build package
-        run: python3 -m build
+        run: .venv/bin/python -m build
         
       - name: Extract version from tag
         id: get_version
         run: echo "VERSION=${GITHUB_REF#refs/tags/v}" >> $GITHUB_OUTPUT
+      - name: Generate changelog
+        id: changelog
+        run: |
+          PREV_TAG=$(git tag --sort=-version:refname | grep -m 1 -v "^${GITHUB_REF#refs/tags/}$")
+          if [ -n "$PREV_TAG" ]; then
+            CHANGELOG=$(git log --pretty=format:"- %s" "${PREV_TAG}..HEAD")
+          else
+            CHANGELOG="Initial release"
+          fi
+          # Write multiline to output
+          {
+            echo "CHANGELOG<<EOF"
+            echo "$CHANGELOG"
+            echo "EOF"
+          } >> $GITHUB_OUTPUT
 
       - name: Upload to Gitea PyPI registry
         env:
           TWINE_USERNAME: ${{ secrets.PYPI_USERNAME }}
           TWINE_PASSWORD: ${{ secrets.PYPI_TOKEN }}
         run: |
-          python3 -m twine upload --repository-url https://git.wrede.ca/api/packages/andreas/pypi dist/*
+          .venv/bin/python3 -m twine upload --repository-url https://git.wrede.ca/api/packages/andreas/pypi dist/*
 
       - name: Create release
         uses: actions/gitea-release-action@v1
@@ -48,4 +60,4 @@ jobs:
             dist/*.whl
             dist/*.tar.gz
           title: "Release ${{ steps.get_version.outputs.VERSION }}"
-          body: "Release version ${{ steps.get_version.outputs.VERSION }}"
+          body: "${{ steps.changelog.outputs.CHANGELOG }}"

.gitignore

@@ -5,6 +5,7 @@ __pycache__/
 *.pyo
 .flake8
 .venv/
+.continue/
 test/
 build/
 dist/
@@ -12,3 +13,6 @@ dist/
 ssl/
 uv.lock
 .hb.yaml
+.superpowers/
+rndc-key
+docs/superpowers/

CHANGELOG.md

@@ -0,0 +1,525 @@
+# Changelog
+
+All notable changes to this project are documented here, organized by release.
+
+## [5.4.0]
+
+### Added
+- gate remote command execution behind allow_remote_command
+- flapping detection suppresses notification storms
+- remove log section from live dashboard (moved to /log)
+- dedicated /log page with journal-backed history and nav item
+- GET /api/0/log — paged, filtered events journal API
+- eventlog writes to dedicated events journal; init/backfill/close wiring
+- events journal read path with filters and backward pagination
+- events journal write primitives (log_event, backfill, get_events_journal)
+- Live Dashboard redesign — 6 columns, hover details, last-alert column
+- extend the record-row design system to Host Overview, Alerts, and About
+- unified record-row redesign of the settings page
+
+### Fixed
+- harden events log — non-dict ring entries, off-loop journal reads, no default-config singleton
+- skip journal scheduling when the event loop is not running
+- dedup log rows between API seed and websocket tail
+- journal startup/shutdown lifecycle events (fired outside journal lifetime)
+- drop typing names not yet used from journal.py import
+- settings toolbar sticks below the site nav instead of scrolling away
+- settings page not scrollable — restore html/body overflow override
+
+---
+
+## [5.3.12]
+
+### Added
+- retire per-channel private flag on profile page
+- settings UI renders per-user edit rights and owner controls
+- settings page accessible to all authenticated users
+- per-user filtering of settings sections
+- scoped non-admin saves through POST /api/0/config
+- threshold form payload carries per-config owner
+- channel ownership rule — owner-presence means private, admin promote/demote
+- scoped threshold-config merge for non-admin saves
+- scoped host merge for non-admin config saves
+- ownership visibility helpers for config entities
+
+### Fixed
+- users loop clobbered requesting username in settings filtering
+- type annotations for mypy parity with master
+- don't re-announce boot/shutdown on SIGHUP restart
+
+---
+
+## [5.3.11]
+
+### Added
+- add Windows hbc client with PyInstaller spec and NSSM install script
+- clear alerts for individual plugin metrics that disappear between samples
+- show alerts for all hosts on Alerts page, not just watched
+
+### Fixed
+- declare plugin interval so stale data waits two full intervals
+- cap event buffer and replay only recent messages on dashboard connect
+- strip plugin timers when pickling Host to prevent save failure
+- correct zero-safe pathconf checks and connectivity prefix match
+- address security vulnerabilities from audit
+- don't purge connectivity/rtt alerts in purge_stale_alerts
+- restore connectivity alerts for overdue/unknown/down hosts on startup
+- clear plugin data and timers on connection UP transition
+- restore host link from Dashboard to Host Overview
+- don't set stale timer until two plugin samples establish real interval
+
+---
+
+## [5.3.10]
+
+### Added
+- clear stale plugin data and persist OAuth users to config
+- auto-scale CPU history graph Y axis
+- add CPU usage history graph to CPU Monitor section
+
+### Fixed
+- remove bak file in bumpminor.sh
+
+---
+
+## [5.3.9]
+
+### Added
+- auto-update CHANGELOG and README in bumpminor.sh
+
+---
+
+## [5.3.8]
+
+### Added
+- Wiki home page with overview and getting started guide
+
+### Fixed
+- Release workflow: use `GITHUB_REF`/`GITHUB_OUTPUT` (Gitea Actions uses GitHub-compatible variable names)
+- Release workflow: replace `head -1` with `grep -m 1` to avoid SIGPIPE (exit 141) in changelog step
+
+---
+
+## [5.3.7]
+
+### Added
+- Dark mode with light/dark/auto theme setting
+- UNKNOWN level filter in Log of Events
+- Per-metric grace period input in threshold settings
+- Replace Dynamic DNS YAML editor with a web form
+- Sort hosts, thresholds, and channels alphabetically on settings page
+- Suppress alerts for unwatched hosts
+
+### Fixed
+- Preserve log message order when replaying history on connect
+
+---
+
+## [5.3.6]
+
+### Added
+- MIT license
+
+### Fixed
+- Correct ZFS pool status threshold operator and add per-metric grace
+- Normalize email and domain fields
+- Move dependencies back under `[project]` in pyproject.toml
+
+---
+
+## [5.3.4]
+
+### Fixed
+- Run full reload after HTTP config publish, not just `config.reload()`
+
+---
+
+## [5.3.3]
+
+### Added
+- Replace YAML threshold editor with a form-based UI
+- Replace multi-select fields with dual-panel picker on settings page
+- Nav bar button to publish pending config changes
+- Host, level, and message filters in Log of Events
+
+### Fixed
+- Remove container max-width; stop stretching inputs on settings page
+
+### Removed
+- Legacy `dyndnshosts`/`drophosts` config keys
+
+---
+
+## [5.3.2]
+
+### Added
+- Retry DNS resolution indefinitely; add `-4`/`-6` address-family flags to `hbc` and `hbc_mini`
+- Replace YAML hosts editor with form-based CRUD table
+- Replace YAML notification channel editor with form-based UI
+
+### Fixed
+- Support list-valued `threshold_config` in hosts table
+- Derive hosts threshold config list from config file keys
+- Replace channel checkboxes in Users table with multi-select
+- Support plugin-level `enabled: false` in threshold config
+- Always populate glance strip for all hosts on page load
+- Fetch host info on initial page load
+
+---
+
+## [5.3.1]
+
+### Added
+- Host info section in Host Overview (fetched and rendered on card expand)
+- `GET /api/0/hosts/{hostname}/info` endpoint
+- Show suffix-matched metric coverage in host info threshold table
+- Move `hbc_version` and `hbc_type` out of `os_info` into the host info section
+
+### Fixed
+- Correct `THRESHOLD_DEFAULTS` metric keys and add missing defaults
+
+---
+
+## [5.3.0]
+
+### Added
+- Profile page self-service: change identity, password, and notification channels
+- Settings page editor with form sections, YAML editors, stage/publish/rollback workflow
+- Config read API: `GET /api/0/config`, `/section/{name}`, `/backups`
+- Config write API: `POST /api/0/config`, `POST /api/0/config/rollback`
+- `configio` module for comment-preserving YAML round-trip writes
+- Multi-provider OAuth2 login page and generic provider routes
+- Log login/logout events to the event log with auth source
+
+### Fixed
+- ZFS monitor alerts dropped on restart with wildcard pool thresholds
+- Preserve OAuth users across config reload
+- Config API error handling, consistent 403 messages, deduplicated key lists
+- Validate password body type; coerce `notification_channels` to strings in profile API
+- Preserve OAuth `client_secret` on roundtrip; harden rollback path validation
+
+---
+
+## [5.2.6]
+
+### Added
+- Alerts host-filter field with URL query parameter and notify URL
+- Optional logo on Gitea OAuth login button
+
+### Fixed
+- Show human-readable duration in re-notification messages
+
+---
+
+## [5.2.5]
+
+### Added
+- Alert CRITICAL on degraded or suspended ZFS pools (ONLINE=OK, DEGRADED=WARNING, all else=CRITICAL)
+- Sign in with Gitea button on login page with OAuth2 redirect/callback routes
+- OAuth2 CSRF state management
+- Host owner shown in glance strip for admin users
+- C port of `hbc_mini` (single-file client in `scripts/c/`)
+
+### Fixed
+- Use `base_url` config for OAuth redirect URI to handle reverse proxy deployments
+- Preserve OAuth users across config reload
+- Escape HTML in login page error display
+
+---
+
+## [5.2.4]
+
+### Added
+- `hbc`/`hbc_mini`: `owner` config field included in `os_info`; server applies to host record
+- Server requests InfoPlugin refresh when a host has no plugin data
+- Event log stores structured dicts; filter by user
+
+### Fixed
+- Strip `_status_code` suffix from displayed metric names in threshold alerts
+- Use plain URL in Mattermost plugin metrics link
+- Fall back to `default_owner` when `os_info` has no owner
+
+---
+
+## [5.2.3]
+
+### Added
+- `hbc`/`hbc_mini`: log name and version at startup
+- Show metric name inline with hostname in alerts and notifications
+
+### Fixed
+- Send shutdown message only if a boot message was previously sent; suppress both on restart
+
+---
+
+## [5.2.2]
+
+### Fixed
+- Retry connection on network error instead of permanently dropping it
+- Silence `aiohttp.access` log; strip plugin prefix in alerts UI
+
+---
+
+## [5.2.1]
+
+### Fixed
+- Threshold and logging improvements
+
+---
+
+## [5.2.0]
+
+### Added
+- `nagios` operator for direct exit-code severity mapping
+
+### Fixed
+- Always show `THRESHOLD_DEFAULTS` in Settings threshold config
+
+---
+
+## [5.1.21]
+
+### Added
+- `nagios_runner` improvements and alerts page fixes
+
+---
+
+## [5.1.20]
+
+### Added
+- Generic threshold matching for `nagios_runner` with `{check_name}` display support
+
+### Fixed
+- Reduce default hysteresis from 10% to 2%
+- Show recovery threshold in alerts UI
+
+---
+
+## [5.1.19]
+
+### Added
+- Exclude ZFS ARC from `memory_percent`
+- Add `uptime_seconds` to `cpu_monitor`
+
+### Fixed
+- Send boot/shutdown message on the first open connection, not blindly on the first in list
+
+---
+
+## [5.1.18]
+
+### Added
+- Fetch-based Update/Delete buttons with toast notifications on Host Overview
+
+### Fixed
+- Settings thresholds show correct per-config metrics; miscellaneous `hbc` fixes
+
+---
+
+## [5.1.17]
+
+### Added
+- Owner Update/Delete buttons on Host Overview; purge stale alerts on reload
+- Retry `AsyncConnection.open()` indefinitely; drop IPv6 only on early startup failure
+- Alert pie chart in the nav bar
+
+### Fixed
+- Make Alerts page scrollable
+
+---
+
+## [5.1.16]
+
+### Added
+- Generic `ping_monitor` thresholds; round RTT to nearest ms
+
+---
+
+## [5.1.15]
+
+### Added
+- Link hostnames in Live Dashboard to Host Overview
+- Threshold Configurations section on settings page
+
+### Fixed
+- Suppress notifications on alert de-escalation (e.g. CRITICAL→WARNING)
+- Suppress recover messages for down durations under 4 seconds
+
+---
+
+## [5.1.14]
+
+### Added
+- ZFS pool renderer in Host Overview
+
+---
+
+## [5.1.13]
+
+### Added
+- ZFS monitor plugin
+- Host-level watch flag to suppress notifications
+- Filter Live Dashboard and Host Overview by owner/manager
+- Composable `threshold_config` list for per-host threshold layering
+- Restart on SIGHUP in `hbc` and `hbc_mini`
+
+### Fixed
+- Mask `api_password` and `access_token` in settings page
+
+---
+
+## [5.1.12]
+
+Internal release — no user-visible changes.
+
+---
+
+## [5.1.11]
+
+### Fixed
+- Install under Docker
+- Clean up install script
+
+---
+
+## [5.1.10]
+
+### Fixed
+- Synchronize version in `hbc_mini`
+- Install script no longer overwrites itself
+
+---
+
+## [5.1.9]
+
+### Added
+- Install `hbc_mini` via package or install script
+
+---
+
+## [5.1.8]
+
+### Added
+- Track `hbc` type and version
+
+### Fixed
+- Nav bar position
+
+---
+
+## [5.1.7]
+
+### Added
+- `hbc_mini`: single-file heartbeat client
+
+### Fixed
+- Drop dead connections on protocol error
+
+---
+
+## [5.1.6]
+
+### Fixed
+- Simplify event log usage; fix argument handling
+
+---
+
+## [5.1.5]
+
+### Added
+- Update `hbc` via `hb_install.sh` instead of code patching
+
+---
+
+## [5.1.4]
+
+### Added
+- Redesign Plugin Metrics page as Host Overview
+
+---
+
+## [5.1.3]
+
+### Added
+- Validate absolute command paths at `nagios_runner` init
+- Async subprocess in `nagios_runner` with stderr capture and signal handling
+- `skip_reason` field on `Plugin`; surface in `PluginLoader` init messaging
+
+### Fixed
+- Use `shlex.split()` for `nagios_runner` path validation to handle quoted paths
+- Reconfigure logging to syslog after `daemonize()`
+
+---
+
+## [5.1.2]
+
+### Fixed
+- Plugin config lookup shadowed by `CLIENT_DEFAULTS` plugins key
+- Apply grace period to all threshold alerts before logging/notifying
+- RECOVER routing: use consistent level name and route via alerted channel
+- Early reminder notifications and lost recovery notifications
+- Non-alerting of overdue hosts
+
+### Added
+- Swiss clock widget in the UI
+
+---
+
+## [5.1.1]
+
+### Added
+- SMS and Matrix notification channels
+- CLI commands `stop`, `restart`, and `reload` for `hbd`
+- WebSocket endpoint at `http://.../ws`
+- Mobile HTML pages
+
+### Fixed
+- Profile not updating
+- Sortable columns in tables
+
+---
+
+## [5.1.0]
+
+### Added
+- Ping monitor plugin
+- Persist state to pickle file; restart timers on server restart
+- SIGHUP config reload for `hbd`
+- Renotify on CRITICAL only; persistent user sessions
+- RTT count threshold
+
+### Fixed
+- Bogus notification on new clients
+- Show "overdue" in alerts instead of null
+
+---
+
+## [5.0.12]
+
+### Added
+- User management and settings page
+
+---
+
+## [5.0.10]
+
+### Added
+- Publish package to Gitea PyPI registry
+
+---
+
+## [5.0.9]
+
+### Added
+- Use `SO_TIMESTAMP` for RTT measurement (Linux, FreeBSD, macOS)
+- Persist state to pickle file; restart timers on restart
+
+---
+
+## [5.0.6]
+
+### Added
+- Major codebase refactoring: restructured into client/server components
+- Per-client threshold configuration
+- Display and acknowledge alerts in the UI
+- Proper `hbc` termination; `hbd` config reloadable at runtime

CLAUDE. md

@@ -1,4 +0,0 @@
-1. Don't assume. Don't hide confusion. Surface tradeoffs.
-2. Minimum code that solves the problem. Nothing speculative.
-3. Touch only what you must. Clean up only your own mess.
-4. Define success criteria. Loop until verified.

CLAUDE.md

@@ -0,0 +1,8 @@
+1. Don't assume. Don't hide confusion. Surface tradeoffs.
+2. Minimum code that solves the problem. Nothing speculative.
+3. Touch only what you must. Clean up only your own mess.
+4. Define success criteria. Loop until verified.
+
+This project's hbd daemon is deployed and in production on host w02 
+under user 'nagios', which you can reach via ssh nagios@w02.
+ The hbc client is used on 30 hosts in multiple networks.

Home.md

@@ -0,0 +1,210 @@
+# Heartbeat
+
+Heartbeat is a lightweight host monitoring system built around a simple idea: each machine you want to monitor runs a small client (`hbc`) that sends a UDP "heartbeat" packet to a central server (`hbd`) on a regular interval. If a heartbeat stops arriving, you get notified. Alongside reachability, clients can ship system metrics — CPU, memory, disk, network — and the server will alert you when any of those cross a threshold.
+
+## How it works
+
+```
+  [ monitored host ]              [ your server ]
+  ┌─────────────┐  UDP 50003     ┌────────────────────────┐
+  │     hbc     │ ────────────>  │  hbd                   │
+  │             │                │  host state tracking   │
+  │  plugins:   │  <──────────── │  threshold alerting    │
+  │  cpu, mem,  │  ACK / CMD     │  notifications         │
+  │  disk, ...  │                │  web dashboard + API   │
+  └─────────────┘                └────────────────────────┘
+```
+
+- **hbd** — the server daemon. Tracks which hosts are alive, evaluates metric thresholds, fires notifications, serves the web dashboard and REST API.
+- **hbc** — the client. Sends heartbeats and plugin data over UDP. Runs on any Linux/BSD/macOS host.
+- **hbc_mini** — a zero-dependency single-file alternative (`hbc_mini.py` or `hbc_mini.c`) for hosts where you can't install Python packages.
+
+Notifications can go to Pushover, email, Mattermost, Matrix, Signal, or VoIP.ms SMS. The dashboard shows host connectivity, RTT graphs, active alerts, and per-host plugin metrics in real time via WebSocket.
+
+---
+
+## Getting started
+
+This tutorial sets up a server on one machine and a client on a second machine. You'll end up with a working dashboard and your first host being monitored.
+
+### 1. Install the server
+
+On the machine that will run `hbd`:
+
+```bash
+git clone https://git.wrede.ca/andreas/heartbeat.git
+cd heartbeat
+python3 -m venv .venv
+source .venv/bin/activate
+pip install .
+```
+
+Verify the install:
+
+```bash
+hbd --help
+```
+
+### 2. Create a server config
+
+Create `~/.hb.yaml`:
+
+```yaml
+hb_port: 50003       # UDP port — clients send heartbeats here
+hbd_port: 50004      # HTTP port — web dashboard and API
+ws_port: 50005       # WebSocket port — live dashboard updates
+
+interval: 20         # Expected heartbeat interval (seconds)
+grace: 2             # Seconds of slack before a host is considered overdue
+
+pickfile: ~/.hb.pick
+pidfile:  ~/.hb.pid
+logfile:  ~/.hb.log
+```
+
+That's enough to get started. No hosts, no users, no notifications needed yet — the server will accept any client that connects.
+
+### 3. Start the server
+
+```bash
+hbd serve -c ~/.hb.yaml -f -v
+```
+
+`-f` keeps it in the foreground so you can watch the log. You should see:
+
+```
+Heartbeat daemon starting on UDP :50003, HTTP :50004, WS :50005
+```
+
+Open `http://your-server:50004/live` in a browser. The dashboard is empty for now.
+
+### 4. Install the client on a host to monitor
+
+On the machine you want to monitor (must be able to reach the server on UDP 50003):
+
+```bash
+pip install hbd      # or: copy scripts/hbc_mini.py if you can't install packages
+```
+
+#### Quick start — no config file
+
+```bash
+hbc your-server.example.com
+```
+
+Within a few seconds the server log will show the host checking in, and it will appear on the dashboard.
+
+#### With a config file
+
+Create `~/.hbc.yaml` on the client host:
+
+```yaml
+hb_port: 50003
+interval: 10         # Send a heartbeat every 10 seconds
+
+plugins:
+  cpu_monitor:
+    interval: 60
+  memory_monitor:
+    interval: 60
+  disk_monitor:
+    interval: 60
+```
+
+Then start the client:
+
+```bash
+hbc -c ~/.hbc.yaml your-server.example.com
+```
+
+Send a boot message at startup so the server logs when the host came up:
+
+```bash
+hbc -b -c ~/.hbc.yaml your-server.example.com
+```
+
+Run as a daemon (logs go to syslog):
+
+```bash
+hbc -d -b -c ~/.hbc.yaml your-server.example.com
+```
+
+### 5. View the dashboard
+
+Open `http://your-server:50004/live`. You'll see the monitored host, its last heartbeat time, and RTT. Click the host name to see plugin metrics.
+
+Navigate to `/plugins/<hostname>` for CPU, memory, and disk graphs.
+
+### 6. Add a notification channel (optional)
+
+Edit `~/.hb.yaml` on the server:
+
+```yaml
+notification_channels:
+  pushover_ops:
+    type: pushover
+    token: YOUR_APP_TOKEN
+    user: YOUR_USER_KEY
+
+users:
+  alice:
+    password: pbkdf2:sha256:...    # generate: hbd passwd alice
+    admin: true
+    notification_channels: [pushover_ops]
+
+default_owner: alice
+```
+
+Generate the password hash:
+
+```bash
+hbd passwd alice
+```
+
+Paste the output into the config, then reload:
+
+```bash
+hbd reload
+```
+
+Test the channel:
+
+```bash
+hbd notify
+```
+
+### 7. Set a threshold alert (optional)
+
+Add to `~/.hb.yaml`:
+
+```yaml
+thresholds:
+  cpu_monitor:
+    cpu_percent:
+      warning: 80.0
+      critical: 90.0
+  disk_monitor:
+    partitions:
+      /:
+        percent:
+          warning: 80.0
+          critical: 90.0
+```
+
+Reload: `hbd reload`. The server will now alert when a monitored host crosses these values.
+
+---
+
+## What's next
+
+| Topic | Where to look |
+|---|---|
+| Full server config reference | [README — Server](https://git.wrede.ca/andreas/heartbeat/src/branch/master/README.md#server-hbd) |
+| Client options and all plugins | [README — Client](https://git.wrede.ca/andreas/heartbeat/src/branch/master/README.md#client-hbc) |
+| Threshold alerting details | [THRESHOLD_ALERTING.md](https://git.wrede.ca/andreas/heartbeat/src/branch/master/docs/THRESHOLD_ALERTING.md) |
+| Notification channels | [NOTIFICATIONS.md](https://git.wrede.ca/andreas/heartbeat/src/branch/master/docs/NOTIFICATIONS.md) |
+| User accounts and roles | [USERS.md](https://git.wrede.ca/andreas/heartbeat/src/branch/master/docs/USERS.md) |
+| Writing a custom plugin | [PLUGIN_DEVELOPMENT.md](https://git.wrede.ca/andreas/heartbeat/src/branch/master/docs/PLUGIN_DEVELOPMENT.md) |
+| Nagios check integration | [NAGIOS_INTEGRATION.md](https://git.wrede.ca/andreas/heartbeat/src/branch/master/docs/NAGIOS_INTEGRATION.md) |
+| REST API | [HTTP_API.md](https://git.wrede.ca/andreas/heartbeat/src/branch/master/docs/HTTP_API.md) |
+| Zero-dependency client | [README — hbc_mini](https://git.wrede.ca/andreas/heartbeat/src/branch/master/README.md#hbc_mini--zero-dependency-client) |

LICENSE.md

@@ -0,0 +1,21 @@
+# MIT License
+
+Copyright (c) 2002 - 2026 Andreas Wrede
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

docs/DARK_MODE.md

@@ -0,0 +1,66 @@
+# Dark Mode
+
+Every page in the Heartbeat web UI supports light mode, dark mode, and automatic (follows the OS/browser setting). Each user picks their preference independently; it is stored in the browser and takes effect immediately without a page reload.
+
+---
+
+## Choosing a theme
+
+Open your profile page (`/profile`) and scroll to the **Appearance** section. Click one of the three buttons:
+
+| Button | Behaviour |
+|--------|-----------|
+| **Auto** | Follows the OS or browser dark-mode preference. Updates live if the system setting changes. |
+| **Light** | Always light, regardless of system setting. |
+| **Dark** | Always dark, regardless of system setting. |
+
+The preference is stored in `localStorage` under the key `hbd_theme` and applies to the current browser only. Clearing browser storage resets it to **Auto**.
+
+---
+
+## Implementation notes
+
+### No flash of unstyled content
+
+A small synchronous `<script>` runs at the very top of `<head>`, before any CSS is parsed, and sets `data-theme="dark"` on `<html>` when the stored preference (or the system setting in auto mode) calls for dark. Because it runs before paint, there is no visible flicker on page load.
+
+### CSS custom properties
+
+All colours are expressed as CSS custom properties defined in `head.html`:
+
+```
+:root                    — light-mode values (default)
+html[data-theme="dark"]  — dark-mode overrides
+```
+
+Key variables:
+
+| Variable | Purpose |
+|----------|---------|
+| `--bg` | Page background |
+| `--surface` | Card / panel background |
+| `--surface-2` / `--surface-3` | Slightly lighter/darker surfaces (table rows, hover states) |
+| `--text` / `--text-sec` / `--text-muted` | Primary, secondary, muted text |
+| `--border` / `--border-2`…`4` | Border shades from prominent to faint |
+| `--link` | Hyperlink and interactive-element colour |
+| `--nav-bg` | Navigation bar background |
+| `--input-bg` / `--input-border` | Form control colours |
+| `--shadow` / `--shadow-sm` | Box-shadow alphas |
+
+A single global rule in `head.html` themes all `<input>`, `<select>`, and `<textarea>` elements across every page at once:
+
+```css
+html[data-theme="dark"] input:not([type=checkbox]):not([type=radio]),
+html[data-theme="dark"] select,
+html[data-theme="dark"] textarea { … }
+```
+
+Each page template adds its own `html[data-theme="dark"]` block for page-specific elements (cards, tables, badges, etc.).
+
+### Auto-mode live updates
+
+A `matchMedia` change listener in `head.html` updates `data-theme` whenever the OS preference changes, so users in **Auto** mode see the theme switch without reloading.
+
+### Semantic colours are unchanged
+
+Alert colours (red for critical, orange for warning, green for ok) and status indicators are intentionally left as fixed values — they are semantic signals, not surface colours, and look correct on both light and dark backgrounds.

docs/HTTP_API.md

@@ -53,6 +53,17 @@ See [User Management](USERS.md) for full authentication documentation.
 |--------|------|-------------|------|
 | `GET` | `/api/0/users` | List all users | Admin |
 | `GET` | `/api/0/users/me` | Own profile | Authenticated |
+| `PUT` | `/api/0/users/me` | Update own profile | Authenticated |
+
+### Notification Channels
+
+| Method | Path | Description | Role |
+|--------|------|-------------|------|
+| `GET` | `/api/0/notification_channel_types` | Channel type schemas | Authenticated |
+| `GET` | `/api/0/notification_channels` | List visible channels | Authenticated |
+| `POST` | `/api/0/notification_channels` | Create a channel | Authenticated |
+| `PUT` | `/api/0/notification_channels/{name}` | Update a channel | Owner or Admin |
+| `DELETE` | `/api/0/notification_channels/{name}` | Delete a channel | Owner or Admin |
 
 ### Host Management
 
@@ -203,6 +214,101 @@ Changes take effect immediately but are not written back to the config file. Upd
 
 ---
 
+---
+
+### Notification Channel Endpoints
+
+Channels are visible to all users by default. Channels marked `private: true` are only visible to their owner. Admins see all channels.
+
+#### GET /api/0/notification_channel_types
+Return the schema for every supported notifier type. Used by the web UI to dynamically render the channel creation form.
+
+**Response:**
+```json
+{
+  "pushover": {
+    "label": "Pushover",
+    "fields": [
+      {"key": "token",  "label": "App token",  "type": "secret", "required": true},
+      {"key": "user",   "label": "User key",   "type": "secret", "required": true},
+      {"key": "sound",  "label": "Sound",      "type": "text",   "required": false}
+    ]
+  },
+  "email": { "label": "E-mail", "fields": [ ... ] },
+  ...
+}
+```
+
+---
+
+#### GET /api/0/notification_channels
+List channels visible to the current user (public channels + own private channels). Admins receive all channels.
+
+**Response:**
+```json
+[
+  {
+    "name": "pushover_ops",
+    "type": "pushover",
+    "type_label": "Pushover",
+    "owner": null,
+    "private": false,
+    "min_level": "WARNING",
+    "fields": [
+      {"key": "token", "label": "App token", "value": "•••", "sensitive": true},
+      {"key": "user",  "label": "User key",  "value": "•••", "sensitive": true}
+    ]
+  }
+]
+```
+
+Sensitive fields (`type: "secret"`) are always returned as `"•••"`.
+
+---
+
+#### POST /api/0/notification_channels
+Create a new channel. The creating user becomes the channel's `owner`.
+
+**Request body:**
+```json
+{
+  "name": "my_pushover",
+  "type": "pushover",
+  "token": "app-token",
+  "user": "user-key",
+  "min_level": "WARNING",
+  "private": true
+}
+```
+
+**Response:** `{"ok": true, "name": "my_pushover"}`
+
+**Status codes:** `200 OK`, `400` (missing required field or unknown type), `409` (name already exists)
+
+---
+
+#### PUT /api/0/notification_channels/{name}
+Update an existing channel. Only the channel owner or an admin may update it.
+
+Secret fields sent as `"•••"` are preserved from the existing config (same pattern as OAuth secrets in the admin config editor).
+
+**Request body:** same shape as POST, `name` ignored (taken from URL).
+
+**Response:** `{"ok": true}`
+
+**Status codes:** `200 OK`, `403 Forbidden`, `404 Not Found`
+
+---
+
+#### DELETE /api/0/notification_channels/{name}
+Delete a channel. Only the channel owner or an admin may delete it.
+
+**Response:** `{"ok": true}`
+
+**Status codes:** `200 OK`, `403 Forbidden`, `404 Not Found`
+
+---
+
 ### Alert Endpoints
 
 #### GET /api/0/hosts/{hostname}/alerts

docs/MESSAGE_JOURNAL.md

@@ -406,6 +406,21 @@ Potential improvements for future versions:
 - Journal file encryption
 - Signed journal entries
 
+## Events journal
+
+Alert/connectivity events shown on the `/log` page are written to a second,
+dedicated journal (one JSON object per line, same rotation mechanism):
+
+| Key | Default | Meaning |
+|---|---|---|
+| `events_journal_file` | `events.journal` | Filename inside `journal_dir` |
+| `events_journal_max_size` | 10 MB | Rotation threshold |
+| `events_journal_max_backups` | 10 | Rotated files kept |
+
+`journal_dir` and `journal_enabled` are shared with the message journal.
+On startup, if the events journal file is empty, it is seeded once from the
+pickled in-memory message ring so history survives the upgrade.
+
 ## See Also
 
 - [Configuration Guide](../hbd/config.py) - Full configuration options

docs/NAGIOS_INTEGRATION.md

@@ -104,11 +104,6 @@ The `nagios_runner` plugin collects:
 - `{name}_{metric}_min` - Minimum value (if present)
 - `{name}_{metric}_max` - Maximum value (if present)
 
-**Overall:**
-- `overall_status` - Worst status from all commands
-- `overall_status_code` - Worst status code
-- `plugin_count` - Number of Nagios plugins executed
-
 ## Configuration Options
 
 ```yaml

docs/NOTIFICATIONS.md

@@ -9,6 +9,7 @@ Notifications are dispatched to the **owner and managers** of a host, each via t
 ```
 Alert event (udp.py / threshold.py)
   └─ notify.send_notification(host_name, Notification)
+       ├─ flap.observe(host, service, level) → pass | trip | suppress
        ├─ look up host.owner + host.managers
        ├─ for each user → user.notification_channels
        └─ for each channel → _dispatch_to_channel (filtered by min_level)
@@ -19,6 +20,7 @@ Every notification carries:
 - **body** — detail message (metric value, threshold, duration)
 - **url** — link to the plugin metrics page (`{base_url}/plugins#{hostname}`)
 - **level** — `RECOVER | WARNING | CRITICAL | INFO`
+- **service** — flap-detection key within the host (empty = the host itself)
 
 ## Configuration
 
@@ -30,9 +32,18 @@ Set `base_url` so notification links point to your hbd instance:
 base_url: https://hbd.example.com
 ```
 
-### Global channel definitions
+### Channel definitions
 
-Define channels once; reference them by name from user configs:
+Channels are defined under `notification_channels`. Each entry specifies a delivery type and its credentials. Ownership is the single visibility signal:
+
+| Field | Default | Description |
+|---|---|---|
+| `owner` | *(absent)* | Owning username. Present = private to that user; absent = global. |
+| `min_level` | `WARNING` | Minimum alert level this channel receives. |
+
+(The former `private` flag is retired; leftover `private` keys are ignored and dropped on the next edit.)
+
+**Global channels** (no `owner`; set in the config file or by an admin) can be selected by all users but edited only by admins:
 
 ```yaml
 notification_channels:
@@ -41,7 +52,7 @@ notification_channels:
     type: pushover
     token: your-app-token
     user: your-user-key
-    min_level: WARNING        # optional, default: WARNING
+    min_level: WARNING
 
   email_ops:
     type: email
@@ -58,14 +69,14 @@ notification_channels:
     homeserver: https://matrix.example.org
     access_token: syt_xxx
     room_id: "!abc:matrix.example.org"
-    min_level: CRITICAL       # only send critical alerts to this room
+    min_level: CRITICAL
 
   sms_oncall:
     type: sms_voipms
     api_user: me@example.com
     api_password: secret
-    did: "5551234567"         # your voip.ms DID number
+    did: "5551234567"
-    dst: "5559876543"         # destination number
+    dst: "5559876543"
     min_level: CRITICAL
 
   signal_ops:
@@ -82,9 +93,31 @@ notification_channels:
     username: heartbeat-bot
 ```
 
+**User-created channels** are written by authenticated users through the API, their profile page, or the settings page. They carry an `owner` field and are private to that user:
+
+```yaml
+notification_channels:
+
+  alice_personal:
+    type: pushover
+    token: personal-token
+    user: personal-key
+    owner: alice          # private to alice
+```
+
+### Channel visibility
+
+| Channel | Who can see / select it | Who can edit it |
+|---|---|---|
+| No `owner` (global) | All users | Admins |
+| `owner` set (private) | Only the `owner` | The owner |
+| Any channel | Admins always see everything | Admins |
+
+Admins can **promote** a private channel to global by clearing its owner on the settings page, or **demote** a global channel by assigning an owner.
+
 ### Users with notification channels
 
-Each user lists which global channels they receive notifications on:
+Each user lists which channels they receive notifications on. Users can manage their own selection from the profile page:
 
 ```yaml
 users:
@@ -237,6 +270,43 @@ min_level: WARNING
 
 Reminder notifications (re-notify) are sent only for CRITICAL level alerts.
 
+## Flap detection
+
+A check that toggles between OK and alerting produces a notification per swing. Flap
+detection silences it after the first few.
+
+A **`(host, service)`** pair is flapping once it exceeds `flap_count` WARNING/CRITICAL
+notifications within `flap_interval` minutes. Threshold alerts key on their metric path,
+so a flapping disk check does not silence an unrelated CPU alert; connectivity, boot and
+shutdown events key on the host itself.
+
+```yaml
+flap_count: 5       # notifications within the window that trip flapping (0 disables)
+flap_interval: 10   # minutes — both the counting window and the quiet window
+```
+
+Lifecycle:
+
+| Event | Effect |
+|---|---|
+| Alerts 1..`flap_count` within the window | Delivered normally |
+| Alert `flap_count + 1` | Delivered with ` Now flapping!! No more messages!` appended to the body |
+| Every notification after that | Dropped — including RECOVER and INFO |
+| RECOVER while flapping | Dropped, and starts the `flap_interval` quiet window |
+| WARNING/CRITICAL during the quiet window | Restarts the quiet window; still flapping |
+| Quiet window elapses | Flapping ends **silently** — no notification |
+
+Only outbound notifications are suppressed. `notify.eventlog` keeps recording every event,
+so the journal and the `/log` page retain the full history of the flap.
+
+Flapping pairs appear in each host's `stateinfo()` under `flapping` (a list of service
+keys; `""` means the host itself) and render as an amber **flapping** badge next to the
+host name on the live dashboard, with the affected services in its tooltip.
+
+State lives at module level in `hbd/server/flap.py` and is never pickled — a server restart
+starts every check with a clean slate. Hosts with `watch: false` never reach the
+notification path, so they never flap.
+
 ## API reference
 
 ### `send_notification(host_name, notif) -> dict`
@@ -270,6 +340,7 @@ Called once at startup from `main.py`. Pass the running asyncio event loop so Ma
 - Check that the host has an `owner` or `managers` set
 - Check that users have `notification_channels` listed
 - Check that the channel names in user config match keys under `notification_channels:`
+- If a user can't select a channel, check whether it has an `owner` other than that user
 
 **min_level filtering too aggressive:**
 - Default is `WARNING` — both WARNING and CRITICAL are sent

docs/PLUGIN_DEVELOPMENT.md

@@ -8,6 +8,7 @@ This guide explains how to create custom plugins for the Heartbeat monitoring sy
 - [Plugin Types](#plugin-types)
 - [Creating a Plugin](#creating-a-plugin)
 - [Plugin Lifecycle](#plugin-lifecycle)
+- [Server-initiated InfoPlugin refresh](#server-initiated-infoplugin-refresh)
 - [Configuration](#configuration)
 - [Best Practices](#best-practices)
 - [Examples](#examples)
@@ -250,6 +251,28 @@ Understanding the plugin lifecycle helps you implement plugins correctly:
    └─> Plugin releases resources, closes connections
 ```
 
+## Server-initiated InfoPlugin refresh
+
+When a heartbeat packet arrives from a host the server has no plugin data for (e.g. after a server restart), the server sets `request_update = 1` in the ACK reply. The client detects this flag and immediately re-runs all InfoPlugins — clearing their cached results first — then resends the data as PLG messages.
+
+This means InfoPlugin data will always reach the server as soon as possible without requiring a client restart. No action is needed from plugin authors: the framework handles cache invalidation and re-collection automatically.
+
+The lifecycle for this case looks like:
+
+```
+Server restarts, host reconnects
+   └─> hbd receives HTB with no existing plugin_data for host
+   └─> hbd sets request_update=1 in ACK
+
+Client receives ACK
+   └─> Detects request_update flag
+   └─> Clears _cache on every registered InfoPlugin
+   └─> Calls collect() on each InfoPlugin
+   └─> Sends fresh PLG messages to server
+```
+
+If you write an `InfoPlugin` with side effects in `_collect_info()` (opening connections, writing files, etc.), be aware it may be called more than once per client session when this mechanism triggers.
+
 ## Configuration
 
 ### Plugin-Specific Configuration

docs/THRESHOLD_ALERTING.md

@@ -256,6 +256,56 @@ disk_monitor:
         operator: "<"
 ```
 
+### ZFS Monitor
+
+ZFS pool health is checked automatically for every pool. A pool in any state
+other than `ONLINE` (e.g. `DEGRADED`, `SUSPENDED`, `FAULTED`, `UNAVAIL`) raises
+a **CRITICAL** alert by default — no configuration required.
+
+The default threshold is equivalent to:
+
+```yaml
+zfs_monitor:
+  pools:
+    '*':
+      status:
+        warning: 1
+        critical: 2
+        operator: ">"
+        hysteresis: 0.0
+        display: "ZFS pool {pool_name} is {health}"
+```
+
+`'*'` matches every pool on the host. The notification message includes the pool
+name and its current health string, e.g. `ZFS pool tank is DEGRADED`.
+
+**Override for specific pools** — named pool entries take priority over `'*'`:
+
+```yaml
+zfs_monitor:
+  pools:
+    # Suppress health alerts for a scratch pool (not mission-critical)
+    scratch:
+      status:
+        enabled: false
+
+    # Capacity threshold for a specific pool
+    tank:
+      capacity:
+        warning: 75.0
+        critical: 90.0
+        operator: ">"
+        hysteresis: 0.05
+```
+
+**Alert state paths** follow the pattern `zfs_monitor.<pool_name>.status`,
+so acknowledgements and silences target individual pools:
+
+```
+zfs_monitor.tank.status
+zfs_monitor.backup.status
+```
+
 ### Network Monitor
 
 ```yaml
@@ -814,34 +864,32 @@ Planned features:
 
 ## Multi-Threshold Configuration
 
-**New in version 2.0**: Support for multiple named threshold configurations with per-host mapping.
+Support for multiple named threshold configurations with per-host mapping and composable layering.
 
 ### Overview
 
 The multi-threshold feature allows you to:
-- Define multiple sets of threshold configurations
+- Define multiple named threshold configurations
-- Map different hosts to different threshold sets
+- Assign one or more configurations to each host
+- Compose configurations by layering — each named config's overrides are applied in order on top of the defaults
 - Use different sensitivity levels for different environments
-- Maintain a default configuration for unmapped hosts
 
 ### Configuration Structure
 
+Named configurations are defined under `threshold_configs`. Each host selects which ones to use via `threshold_config` in the `hosts` section (a string for a single config, or a list to layer multiple):
+
 ```yaml
-# Optional: Set the default configuration name (defaults to "default")
+# Optional: set the default configuration name (defaults to "default")
 default_threshold_config: "default"
 
-# Define multiple named threshold configurations
 threshold_configs:
-  # Configuration name 1
   default:
     thresholds:
-      # Standard threshold definitions
       cpu_monitor:
         cpu_percent:
           warning: 80.0
           critical: 90.0
 
-  # Configuration name 2
   high_sensitivity:
     thresholds:
       cpu_monitor:
@@ -849,7 +897,6 @@ threshold_configs:
           warning: 60.0
           critical: 75.0
 
-  # Configuration name 3
   low_sensitivity:
     thresholds:
       cpu_monitor:
@@ -857,14 +904,77 @@ threshold_configs:
           warning: 90.0
           critical: 95.0
 
-# Map specific hosts to specific configurations
+hosts:
-host_threshold_mapping:
+  prod-web-01:
-  prod-web-01: high_sensitivity
+    threshold_config: high_sensitivity   # single config
-  prod-web-02: high_sensitivity
+
-  dev-server-01: low_sensitivity
+  dev-server-01:
-  # Unmapped hosts use default_threshold_config
+    threshold_config: low_sensitivity
+
+  # Hosts with no threshold_config use default_threshold_config
 ```
 
+### Composable Configurations (list form)
+
+`threshold_config` can be a list. Configs are applied **left to right**: the defaults are the base, then each named config's overrides are layered on top. Later entries in the list win on any metric they define.
+
+```yaml
+threshold_configs:
+  default:
+    thresholds:
+      cpu_monitor:
+        cpu_percent: {warning: 80, critical: 90}
+      memory_monitor:
+        memory_percent: {warning: 85, critical: 95}
+      disk_monitor:
+        partitions:
+          /:
+            percent: {warning: 80, critical: 90}
+
+  # Tighter CPU limits for busy servers
+  high_cpu_load:
+    thresholds:
+      cpu_monitor:
+        cpu_percent: {warning: 60, critical: 75}
+
+  # Tighter disk limits for data-heavy servers
+  busy_disk:
+    thresholds:
+      disk_monitor:
+        partitions:
+          /:
+            percent: {warning: 70, critical: 85}
+
+hosts:
+  # Gets default thresholds only
+  web-01:
+    threshold_config: default
+
+  # Gets tighter CPU limits, default memory and disk
+  build-server:
+    threshold_config: high_cpu_load
+
+  # Layers both: tighter CPU AND tighter disk, default memory
+  db-01:
+    threshold_config: [high_cpu_load, busy_disk]
+
+  # Three layers: busy_disk overrides high_cpu_load if they conflict
+  storage-01:
+    threshold_config: [default, high_cpu_load, busy_disk]
+```
+
+**How layering works:**
+
+Starting from the `default` thresholds:
+
+| Layer | Applied config | Effect |
+|-------|---------------|--------|
+| Base  | `default` | all default thresholds |
+| +1    | `high_cpu_load` | cpu_percent overridden to 60/75 |
+| +2    | `busy_disk` | disk percent overridden to 70/85; cpu_percent stays at 60/75 |
+
+Each named config only overrides the metrics it explicitly defines. Metrics not mentioned in a config inherit from the layers beneath.
+
 ### Use Cases
 
 #### 1. Environment-Based Thresholds
@@ -887,11 +997,15 @@ threshold_configs:
           warning: 90.0   # More relaxed for dev
           critical: 98.0
 
-host_threshold_mapping:
+hosts:
-  prod-web-01: production
+  prod-web-01:
-  prod-web-02: production
+    threshold_config: production
-  dev-web-01: development
+  prod-web-02:
-  dev-web-02: development
+    threshold_config: production
+  dev-web-01:
+    threshold_config: development
+  dev-web-02:
+    threshold_config: development
 ```
 
 #### 2. Server Role-Based Thresholds
@@ -914,7 +1028,7 @@ threshold_configs:
           warning: 70.0
           critical: 85.0
       memory_monitor:
-        percent:
+        memory_percent:
           warning: 90.0   # Databases can use high memory
           critical: 97.0
       disk_monitor:
@@ -927,17 +1041,23 @@ threshold_configs:
   cache:
     thresholds:
       memory_monitor:
-        percent:
+        memory_percent:
           warning: 95.0   # Redis/Memcached can use very high memory
           critical: 99.0
 
-host_threshold_mapping:
+hosts:
-  web-01: webserver
+  web-01:
-  web-02: webserver
+    threshold_config: webserver
-  db-01: database
+  web-02:
-  db-02: database
+    threshold_config: webserver
-  redis-01: cache
+  db-01:
-  memcached-01: cache
+    threshold_config: database
+  db-02:
+    threshold_config: database
+  redis-01:
+    threshold_config: cache
+  memcached-01:
+    threshold_config: cache
 ```
 
 #### 3. Sensitivity Levels
@@ -952,7 +1072,7 @@ threshold_configs:
         partitions:
           /:
             percent:
-              warning: 70.0    # Very sensitive
+              warning: 70.0
               critical: 80.0
               hysteresis: 0.15
 
@@ -976,52 +1096,91 @@ threshold_configs:
               critical: 98.0
               hysteresis: 0.05
 
-host_threshold_mapping:
+hosts:
-  payment-gateway: critical
+  payment-gateway:
-  auth-server: critical
+    threshold_config: critical
-  web-01: standard
+  auth-server:
-  web-02: standard
+    threshold_config: critical
-  test-server: relaxed
+  web-01:
+    threshold_config: standard
+  web-02:
+    threshold_config: standard
+  test-server:
+    threshold_config: relaxed
 ```
 
-### Backward Compatibility
+#### 4. Composable Profiles
 
-The legacy single threshold configuration is fully supported:
+Build host-specific thresholds by combining small, focused configs:
 
 ```yaml
-# Old format - still works
-thresholds:
-  cpu_monitor:
-    cpu_percent:
-      warning: 80.0
-      critical: 90.0
-```
-
-This is equivalent to:
-
-```yaml
-# New format
 threshold_configs:
+  # Baseline — everything at default levels
   default:
     thresholds:
       cpu_monitor:
-        cpu_percent:
+        cpu_percent: {warning: 80, critical: 90}
-          warning: 80.0
+      memory_monitor:
-          critical: 90.0
+        memory_percent: {warning: 85, critical: 95}
-```
 
+  # Overlay: tighter CPU only
+  tight_cpu:
+    thresholds:
+      cpu_monitor:
+        cpu_percent: {warning: 60, critical: 75}
+
+  # Overlay: tighter memory only
+  tight_memory:
+    thresholds:
+      memory_monitor:
+        memory_percent: {warning: 70, critical: 85}
+
+  # Overlay: extra disk partition for database servers
+  db_disk:
+    thresholds:
+      disk_monitor:
+        partitions:
+          /var/lib/postgresql:
+            percent: {warning: 75, critical: 88}
+
+hosts:
+  # Plain web server
+  web-01:
+    threshold_config: default
+
+  # Build server: tight CPU, default memory and disk
+  build-01:
+    threshold_config: tight_cpu
+
+  # Database: tight CPU + tight memory + extra disk partition
+  db-01:
+    threshold_config: [tight_cpu, tight_memory, db_disk]
+
+  # Replica database: tight memory + extra disk, normal CPU
+  db-02:
+    threshold_config: [tight_memory, db_disk]
+```
 ### Configuration Priority
 
-1. **Host-specific mapping**: If host is in `host_threshold_mapping`, use that config
+1. **Host `threshold_config` (list)**: Layer each named config's overrides left-to-right on top of the defaults
-2. **Default config**: Use `default_threshold_config` 
+2. **Host `threshold_config` (string)**: Use that single named config directly
-3. **First alphabetically**: If default not found, use first config alphabetically
+3. **`host_threshold_mapping`** (legacy): Same as above, string only
-4. **Legacy fallback**: If `threshold_configs` not present, use `thresholds`
+4. **`default_threshold_config`**: Used for hosts with no mapping
+5. **First alphabetically**: If the default config is not found, use the first config alphabetically
+6. **Legacy `thresholds` section**: Used when `threshold_configs` is absent entirely
 
-### Example: Complete Multi-Threshold Setup
+### Backward Compatibility
 
-See `hbd/config_multi_threshold_example.yaml` for a complete example with:
+The legacy `host_threshold_mapping` top-level key and the flat `thresholds` section are still fully supported:
-- 4 named configurations (default, high_sensitivity, low_sensitivity, database)
+
-- Host-to-config mappings for production, development, and test systems
+```yaml
-- Specialized database server thresholds
+# Still works — equivalent to hosts: {prod-web-01: {threshold_config: high_sensitivity}}
-- Custom display messages with plugin data
+host_threshold_mapping:
+  prod-web-01: high_sensitivity
+
+# Still works — equivalent to threshold_configs: {default: {thresholds: ...}}
+thresholds:
+  cpu_monitor:
+    cpu_percent: {warning: 80, critical: 90}
+```
 

docs/USERS.md

@@ -19,6 +19,15 @@ Users are defined in the server config file. Each host can have an **owner**, ze
 
 `admin` is a flag on the user, not a per-host role. An admin user has owner-level access on every host without being listed as owner/manager/monitor.
 
+### Settings page access
+
+All authenticated users may open `/settings`. Admins see every section; non-admins see only **Notification Channels**, **Hosts**, and **Threshold Configurations**, filtered to global items plus what they own or manage:
+
+- **Owners** may add hosts, delete their hosts, edit host settings, and change access lists (managers/monitors, ownership transfer).
+- **Managers** may edit host settings (watch, dyndns, channel and threshold assignments) but not access lists, and may not delete hosts.
+- Anyone may create private notification channels and threshold configs (owned by them) and assign them — or global ones — to their hosts.
+- Admins promote a private channel/threshold config to global by clearing its owner, or demote by assigning one.
+
 ---
 
 ## Configuration
@@ -36,7 +45,7 @@ users:
   bob:
     full_name: Bob Smith
     password: pbkdf2:sha256:...
-    notification_channels: [pushover_standard]
+    notification_channels: [pushover_standard]   # channels bob has selected
 
   carol:
     full_name: Carol Jones
@@ -46,6 +55,24 @@ default_owner: andreas            # owns hosts with no explicit owner
                                   # falls back to the first admin user if omitted
 ```
 
+### Client-declared host ownership
+
+A host can declare its own owner directly in the hbc or hbc_mini client configuration. This is useful for hosts that are not listed in the server config, or during initial setup before a server-side config entry has been created.
+
+**`~/.hbc.yaml`** (hbc):
+```yaml
+owner: andreas
+```
+
+**`~/.hbc.json`** (hbc_mini):
+```json
+{ "owner": "andreas" }
+```
+
+When set, the value is included in the `os_info` plugin data sent to the server. The server applies it as `host.owner` the first time `os_info` arrives, provided no owner has been configured server-side for that host. Server-configured ownership always takes precedence.
+
+---
+
 ### Assigning roles to hosts
 
 ```yaml
@@ -170,6 +197,32 @@ Return the currently authenticated user's profile.
 
 ---
 
+#### PUT /api/0/users/me
+Update the current user's profile. All fields are optional — send only what you want to change.
+
+**Update display name and avatar:**
+```json
+{ "full_name": "Carol Jones", "avatar": "/avatars/carol.png" }
+```
+
+**Change notification channel selection:**
+```json
+{ "notification_channels": ["pushover_ops", "email_ops"] }
+```
+Only channels visible to the user (global + own) are accepted; others are silently dropped.
+
+**Change password:**
+```json
+{ "password": { "current": "oldpass", "new": "newpass" } }
+```
+Requires the correct current password. New password is hashed before storage.
+
+**Response:** `{"ok": true}`
+
+**Status codes:** `200 OK`, `400` (missing/invalid field), `401` (unauthenticated), `403` (wrong current password)
+
+---
+
 ### Host Access
 
 #### GET /api/0/hosts/{hostname}/access

docs/superpowers/plans/2026-04-25-plugin-error-checking.md

@@ -1,602 +0,0 @@
-# Plugin Error Checking Implementation Plan
-
-> **For agentic workers:** REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (`- [ ]`) syntax for tracking.
-
-**Goal:** Improve plugin error checking in hbc, especially for nagios_runner, and fix logger messages silently discarded in daemon mode.
-
-**Architecture:** Three focused changes across three files: (1) `hbd/client/plugin.py` gains a `skip_reason` attribute on Plugin and updated PluginLoader messaging; (2) `hbd/client/plugins/nagios_runner.py` gains async subprocess execution, stderr capture, signal-killed process handling, and init-time command path validation; (3) `hbd/client/main.py` gains proper post-fork logging reconfiguration to syslog.
-
-**Tech Stack:** Python 3.11+, asyncio, `logging.handlers.SysLogHandler`, pytest
-
----
-
-## File Map
-
-| Action | Path | What changes |
-|---|---|---|
-| Modify | `hbd/client/plugin.py` | `Plugin.__init__` gains `skip_reason`; `PluginLoader` checks it |
-| Modify | `hbd/client/plugins/nagios_runner.py` | async subprocess, stderr, signal codes, init validation, `skip_reason` |
-| Modify | `hbd/client/main.py` | `_reconfigure_logging_for_daemon()` helper; remove redundant syslog calls |
-| Create | `tests/test_plugin.py` | PluginLoader messaging tests |
-| Create | `tests/test_nagios_runner.py` | NagiosRunnerPlugin behaviour tests |
-
-Run tests throughout with:
-```bash
-python -m pytest tests/test_plugin.py tests/test_nagios_runner.py -v
-```
-
----
-
-## Task 1: Plugin.skip_reason + PluginLoader messaging
-
-**Files:**
-- Modify: `hbd/client/plugin.py:40-48` (Plugin.__init__)
-- Modify: `hbd/client/plugin.py:369-381` (PluginLoader.load_from_directory)
-- Create: `tests/test_plugin.py`
-
-- [ ] **Step 1: Write failing tests**
-
-Create `tests/test_plugin.py`:
-
-```python
-import asyncio
-import logging
-import textwrap
-
-from hbd.client.plugin import Plugin, PluginLoader, PluginRegistry
-
-
-def test_plugin_skip_reason_defaults_none(tmp_path):
-    plugin_code = textwrap.dedent("""
-        from hbd.client.plugin import MonitorPlugin
-
-        class MinimalPlugin(MonitorPlugin):
-            name = "minimal"
-            version = "1.0.0"
-            interval = 60
-
-            async def initialize(self):
-                return True
-
-            async def _collect_metrics(self):
-                return {}
-    """)
-    (tmp_path / "minimal.py").write_text(plugin_code)
-    registry = PluginRegistry()
-    loader = PluginLoader(registry)
-    asyncio.run(loader.load_from_directory(tmp_path))
-    plugin = registry.get("minimal")
-    assert plugin is not None
-    assert plugin.skip_reason is None
-
-
-def test_loader_logs_info_when_skip_reason_set(tmp_path, caplog):
-    plugin_code = textwrap.dedent("""
-        from hbd.client.plugin import MonitorPlugin
-
-        class SkippablePlugin(MonitorPlugin):
-            name = "skippable"
-            version = "1.0.0"
-            interval = 60
-
-            async def initialize(self):
-                self.skip_reason = "not configured in yaml"
-                return False
-
-            async def _collect_metrics(self):
-                return {}
-    """)
-    (tmp_path / "skippable.py").write_text(plugin_code)
-    registry = PluginRegistry()
-    loader = PluginLoader(registry)
-
-    with caplog.at_level(logging.INFO, logger="plugin.loader"):
-        count = asyncio.run(loader.load_from_directory(tmp_path))
-
-    assert count == 0
-    assert any("skipped: not configured in yaml" in r.message for r in caplog.records)
-    assert not any("failed initialization" in r.message for r in caplog.records)
-
-
-def test_loader_logs_warning_when_no_skip_reason(tmp_path, caplog):
-    plugin_code = textwrap.dedent("""
-        from hbd.client.plugin import MonitorPlugin
-
-        class FailPlugin(MonitorPlugin):
-            name = "fail"
-            version = "1.0.0"
-            interval = 60
-
-            async def initialize(self):
-                return False
-
-            async def _collect_metrics(self):
-                return {}
-    """)
-    (tmp_path / "fail_plugin.py").write_text(plugin_code)
-    registry = PluginRegistry()
-    loader = PluginLoader(registry)
-
-    with caplog.at_level(logging.WARNING, logger="plugin.loader"):
-        count = asyncio.run(loader.load_from_directory(tmp_path))
-
-    assert count == 0
-    assert any("failed initialization" in r.message for r in caplog.records)
-```
-
-- [ ] **Step 2: Run tests to verify they fail**
-
-```bash
-python -m pytest tests/test_plugin.py -v
-```
-Expected: `test_plugin_skip_reason_defaults_none` FAILS (attribute missing), others may error.
-
-- [ ] **Step 3: Add `skip_reason` to `Plugin.__init__`**
-
-In `hbd/client/plugin.py`, in `Plugin.__init__` (around line 46), add one line:
-
-```python
-def __init__(self, config: Optional[Dict[str, Any]] = None):
-    self.config = config or {}
-    self.logger = logging.getLogger(f"plugin.{self.name}")
-    self._initialized = False
-    self.skip_reason: Optional[str] = None
-```
-
-- [ ] **Step 4: Update PluginLoader messaging**
-
-In `hbd/client/plugin.py`, replace the `if not initialized:` block (around line 372):
-
-```python
-                    if not initialized:
-                        if plugin.skip_reason:
-                            self.logger.info(
-                                f"Plugin {plugin.name} skipped: {plugin.skip_reason}"
-                            )
-                        else:
-                            self.logger.warning(
-                                f"Plugin {plugin.name} failed initialization, skipping"
-                            )
-                        continue
-```
-
-- [ ] **Step 5: Run tests to verify they pass**
-
-```bash
-python -m pytest tests/test_plugin.py -v
-```
-Expected: all 3 tests PASS.
-
-- [ ] **Step 6: Commit**
-
-```bash
-git add hbd/client/plugin.py tests/test_plugin.py
-git commit -m "feat: add skip_reason to Plugin; improve PluginLoader init messaging"
-```
-
----
-
-## Task 2: NagiosRunnerPlugin — skip_reason when no commands
-
-**Files:**
-- Modify: `hbd/client/plugins/nagios_runner.py:88-105` (initialize)
-- Modify: `tests/test_nagios_runner.py` (create)
-
-- [ ] **Step 1: Write failing test**
-
-Create `tests/test_nagios_runner.py`:
-
-```python
-import asyncio
-import logging
-import os
-import stat
-
-import pytest
-
-from hbd.client.plugins.nagios_runner import (
-    NagiosRunnerPlugin,
-    NAGIOS_OK,
-    NAGIOS_WARNING,
-    NAGIOS_CRITICAL,
-    NAGIOS_UNKNOWN,
-)
-
-
-def test_no_commands_sets_skip_reason():
-    plugin = NagiosRunnerPlugin(config={"commands": []})
-    result = asyncio.run(plugin.initialize())
-    assert result is False
-    assert plugin.skip_reason is not None
-    assert "nagios_runner.commands" in plugin.skip_reason
-```
-
-- [ ] **Step 2: Run test to verify it fails**
-
-```bash
-python -m pytest tests/test_nagios_runner.py::test_no_commands_sets_skip_reason -v
-```
-Expected: FAIL — `plugin.skip_reason` is `None`.
-
-- [ ] **Step 3: Set skip_reason in NagiosRunnerPlugin.initialize()**
-
-In `hbd/client/plugins/nagios_runner.py`, replace the early-return block in `initialize()` (around line 96):
-
-```python
-        if not self.commands:
-            self.skip_reason = "no commands configured (add nagios_runner.commands to config)"
-            self.logger.info("No Nagios commands configured")
-            return False
-```
-
-- [ ] **Step 4: Run test to verify it passes**
-
-```bash
-python -m pytest tests/test_nagios_runner.py::test_no_commands_sets_skip_reason -v
-```
-Expected: PASS.
-
-- [ ] **Step 5: Commit**
-
-```bash
-git add hbd/client/plugins/nagios_runner.py tests/test_nagios_runner.py
-git commit -m "feat: set skip_reason on nagios_runner when no commands configured"
-```
-
----
-
-## Task 3: NagiosRunnerPlugin — async subprocess, stderr capture, negative return codes
-
-**Files:**
-- Modify: `hbd/client/plugins/nagios_runner.py` (imports + `_run_nagios_plugin`)
-- Modify: `tests/test_nagios_runner.py`
-
-- [ ] **Step 1: Write failing tests**
-
-Append to `tests/test_nagios_runner.py`:
-
-```python
-def test_stderr_used_when_stdout_empty(tmp_path):
-    script = tmp_path / "check_err.sh"
-    script.write_text("#!/bin/sh\necho 'error from stderr' >&2\nexit 2\n")
-    script.chmod(script.stat().st_mode | stat.S_IEXEC)
-
-    config = {"commands": [{"name": "t", "command": str(script)}], "timeout": 5}
-    plugin = NagiosRunnerPlugin(config=config)
-    asyncio.run(plugin.initialize())
-    data = asyncio.run(plugin._collect_metrics())
-
-    assert "error from stderr" in data["t_output"]
-    assert data["t_status_code"] == NAGIOS_CRITICAL
-
-
-def test_stderr_appended_when_both_present(tmp_path):
-    script = tmp_path / "check_both.sh"
-    script.write_text("#!/bin/sh\necho 'OK - all good'\necho 'extra detail' >&2\nexit 0\n")
-    script.chmod(script.stat().st_mode | stat.S_IEXEC)
-
-    config = {"commands": [{"name": "t", "command": str(script)}], "timeout": 5}
-    plugin = NagiosRunnerPlugin(config=config)
-    asyncio.run(plugin.initialize())
-    data = asyncio.run(plugin._collect_metrics())
-
-    assert "OK - all good" in data["t_output"]
-    assert "extra detail" in data["t_output"]
-    assert data["t_status_code"] == NAGIOS_OK
-
-
-def test_negative_returncode_maps_to_unknown():
-    # kill -9 $$ kills the shell itself; asyncio sees returncode -9
-    config = {"commands": [{"name": "t", "command": "kill -9 $$"}], "timeout": 5}
-    plugin = NagiosRunnerPlugin(config=config)
-    asyncio.run(plugin.initialize())
-    data = asyncio.run(plugin._collect_metrics())
-
-    assert data["t_status_code"] == NAGIOS_UNKNOWN
-    assert "signal" in data["t_output"].lower()
-```
-
-- [ ] **Step 2: Run tests to verify they fail**
-
-```bash
-python -m pytest tests/test_nagios_runner.py::test_stderr_used_when_stdout_empty \
-    tests/test_nagios_runner.py::test_stderr_appended_when_both_present \
-    tests/test_nagios_runner.py::test_negative_returncode_maps_to_unknown -v
-```
-Expected: all FAIL — current implementation ignores stderr and doesn't handle negative codes.
-
-- [ ] **Step 3: Update imports in nagios_runner.py**
-
-Replace the import block at the top of `hbd/client/plugins/nagios_runner.py`:
-
-```python
-import asyncio
-import os
-import re
-from typing import Any, Dict, List, Optional, Tuple
-
-from hbd.client.plugin import MonitorPlugin
-```
-
-(Remove `import subprocess`; add `import asyncio` and `import os`.)
-
-- [ ] **Step 4: Upgrade collection log level from DEBUG to INFO**
-
-In `hbd/client/plugins/nagios_runner.py`, in `_collect_metrics()`, change the debug log (around line 144) so results are visible at INFO level:
-
-```python
-                self.logger.info(
-                    f"Executed {name}: {STATUS_NAMES.get(status_code, 'UNKNOWN')} - {output[:50]}"
-                )
-```
-
-- [ ] **Step 5: Replace `_run_nagios_plugin` with async implementation**
-
-Replace the entire `_run_nagios_plugin` method in `hbd/client/plugins/nagios_runner.py`:
-
-```python
-    async def _run_nagios_plugin(
-        self,
-        command: str
-    ) -> Tuple[int, str, Dict[str, Any]]:
-        """Execute a Nagios plugin and parse its output."""
-        try:
-            proc = await asyncio.create_subprocess_shell(
-                command,
-                stdout=asyncio.subprocess.PIPE,
-                stderr=asyncio.subprocess.PIPE,
-            )
-            try:
-                stdout_bytes, stderr_bytes = await asyncio.wait_for(
-                    proc.communicate(), timeout=self.timeout
-                )
-            except asyncio.TimeoutError:
-                proc.kill()
-                await proc.communicate()
-                self.logger.error(f"Command timed out: {command}")
-                return NAGIOS_UNKNOWN, f"Command timed out after {self.timeout}s", {}
-
-            status_code = proc.returncode
-
-            if status_code < 0:
-                return NAGIOS_UNKNOWN, f"Process killed by signal {-status_code}", {}
-
-            if status_code > 3:
-                status_code = NAGIOS_UNKNOWN
-
-            stdout = stdout_bytes.decode(errors="replace").strip()
-            stderr = stderr_bytes.decode(errors="replace").strip()
-
-            # Parse perfdata from stdout before mixing in stderr
-            perfdata = self._parse_perfdata(stdout)
-
-            # Build status message
-            status_part = stdout.split('|')[0].strip() if '|' in stdout else stdout
-
-            if not stdout and stderr:
-                output_msg = stderr
-            elif stdout and stderr:
-                output_msg = f"{status_part} [stderr: {stderr}]"
-            else:
-                output_msg = status_part
-
-            return status_code, output_msg, perfdata
-
-        except Exception as e:
-            self.logger.error(f"Error executing command: {e}")
-            return NAGIOS_UNKNOWN, f"Execution error: {str(e)}", {}
-```
-
-Also remove the now-unused `self.shell` line from `__init__` (the `shell` config key is no longer used since `create_subprocess_shell` always uses a shell):
-
-In `NagiosRunnerPlugin.__init__`, remove:
-```python
-        self.shell: bool = config.get("shell", True) if config else True
-```
-
-- [ ] **Step 6: Run tests to verify they pass**
-
-```bash
-python -m pytest tests/test_nagios_runner.py -v
-```
-Expected: all tests PASS including the 3 new ones.
-
-- [ ] **Step 7: Commit**
-
-```bash
-git add hbd/client/plugins/nagios_runner.py tests/test_nagios_runner.py
-git commit -m "feat: async subprocess in nagios_runner with stderr capture and signal handling"
-```
-
----
-
-## Task 4: NagiosRunnerPlugin — command path validation at init
-
-**Files:**
-- Modify: `hbd/client/plugins/nagios_runner.py` (initialize)
-- Modify: `tests/test_nagios_runner.py`
-
-- [ ] **Step 1: Write failing tests**
-
-Append to `tests/test_nagios_runner.py`:
-
-```python
-def test_absolute_path_not_found_warns(caplog):
-    fake_cmd = "/nonexistent_hbc_test_path/check_something"
-    config = {"commands": [{"name": "t", "command": fake_cmd}]}
-    plugin = NagiosRunnerPlugin(config=config)
-
-    with caplog.at_level(logging.WARNING, logger="plugin.nagios_runner"):
-        asyncio.run(plugin.initialize())
-
-    assert any("not found" in r.message for r in caplog.records)
-
-
-def test_absolute_path_not_executable_warns(caplog, tmp_path):
-    non_exec = tmp_path / "check_test"
-    non_exec.write_text("#!/bin/sh\necho OK\n")
-    non_exec.chmod(0o644)  # readable but not executable
-
-    config = {"commands": [{"name": "t", "command": str(non_exec)}]}
-    plugin = NagiosRunnerPlugin(config=config)
-
-    with caplog.at_level(logging.WARNING, logger="plugin.nagios_runner"):
-        asyncio.run(plugin.initialize())
-
-    assert any("not executable" in r.message for r in caplog.records)
-
-
-def test_relative_path_not_checked(caplog):
-    # Relative paths (resolved via PATH) must not generate warnings
-    config = {"commands": [{"name": "t", "command": "echo OK"}]}
-    plugin = NagiosRunnerPlugin(config=config)
-
-    with caplog.at_level(logging.WARNING, logger="plugin.nagios_runner"):
-        asyncio.run(plugin.initialize())
-
-    assert not any(
-        "not found" in r.message or "not executable" in r.message
-        for r in caplog.records
-    )
-```
-
-- [ ] **Step 2: Run tests to verify they fail**
-
-```bash
-python -m pytest tests/test_nagios_runner.py::test_absolute_path_not_found_warns \
-    tests/test_nagios_runner.py::test_absolute_path_not_executable_warns \
-    tests/test_nagios_runner.py::test_relative_path_not_checked -v
-```
-Expected: `test_absolute_path_not_found_warns` and `test_absolute_path_not_executable_warns` FAIL (no warnings logged); `test_relative_path_not_checked` may pass.
-
-- [ ] **Step 3: Add command path validation to `initialize()`**
-
-In `hbd/client/plugins/nagios_runner.py`, extend `initialize()` by adding validation after the existing "log each command" loop (after line 103, before `return True`):
-
-```python
-        # Validate absolute command paths early
-        for cmd_config in self.commands:
-            name = cmd_config.get("name", "unnamed")
-            command = cmd_config.get("command", "")
-            if not command:
-                continue
-            exe = command.split()[0]
-            if os.path.isabs(exe):
-                if not os.path.isfile(exe):
-                    self.logger.warning(
-                        f"Command '{name}': executable not found: {exe}"
-                    )
-                elif not os.access(exe, os.X_OK):
-                    self.logger.warning(
-                        f"Command '{name}': executable not executable: {exe}"
-                    )
-```
-
-- [ ] **Step 4: Run full test suite to verify all pass**
-
-```bash
-python -m pytest tests/test_plugin.py tests/test_nagios_runner.py -v
-```
-Expected: all tests PASS.
-
-- [ ] **Step 5: Commit**
-
-```bash
-git add hbd/client/plugins/nagios_runner.py tests/test_nagios_runner.py
-git commit -m "feat: validate absolute command paths at nagios_runner init"
-```
-
----
-
-## Task 5: Daemon mode logging — route to syslog after fork
-
-**Files:**
-- Modify: `hbd/client/main.py` (new helper + updated daemon block)
-
-No automated test for daemonization itself (fork behaviour is hard to unit-test). Manual verification steps are provided below.
-
-- [ ] **Step 1: Add `_reconfigure_logging_for_daemon` helper**
-
-In `hbd/client/main.py`, add this function just before `def build_parser()` (around line 589):
-
-```python
-def _reconfigure_logging_for_daemon(log_level: int) -> None:
-    """Replace StreamHandlers (now writing to /dev/null) with a SysLogHandler."""
-    from logging.handlers import SysLogHandler
-
-    root = logging.getLogger()
-    for handler in root.handlers[:]:
-        root.removeHandler(handler)
-        handler.close()
-
-    try:
-        syslog_handler = SysLogHandler(
-            address="/dev/log",
-            facility=SysLogHandler.LOG_DAEMON,
-        )
-    except OSError:
-        syslog_handler = SysLogHandler(
-            address=("localhost", 514),
-            facility=SysLogHandler.LOG_DAEMON,
-        )
-        # Attach the fallback first so the warning reaches syslog
-        syslog_handler.setFormatter(
-            logging.Formatter("hbc[%(process)d]: %(name)s %(levelname)s: %(message)s")
-        )
-        root.addHandler(syslog_handler)
-        root.setLevel(log_level)
-        logging.warning("/dev/log not found, using syslog UDP localhost:514")
-        return
-
-    syslog_handler.setFormatter(
-        logging.Formatter("hbc[%(process)d]: %(name)s %(levelname)s: %(message)s")
-    )
-    root.addHandler(syslog_handler)
-    root.setLevel(log_level)
-```
-
-- [ ] **Step 2: Update the daemon block in `main()`**
-
-In `hbd/client/main.py`, replace the entire `if args.daemon:` block (lines 664–675):
-
-```python
-    if args.daemon:
-        print("Daemonizing...")
-        daemonize()
-        _reconfigure_logging_for_daemon(log_level)
-        logging.info(f"hbc starting, sending heartbeat to {', '.join(args.hosts)}")
-```
-
-This removes the `import syslog`, `syslog.openlog()`, and `syslog.syslog()` calls (now handled by the logging system) and removes the no-op second `logging.basicConfig()` call.
-
-- [ ] **Step 3: Run existing test suite to confirm no regressions**
-
-```bash
-python -m pytest tests/test_plugin.py tests/test_nagios_runner.py -v
-```
-Expected: all tests still PASS.
-
-- [ ] **Step 4: Manual smoke test — verify syslog output in daemon mode**
-
-```bash
-# In one terminal, tail syslog
-sudo journalctl -f -t hbc
-
-# In another terminal, start hbc in daemon mode (replace HOST with a real or dummy host)
-python -m hbd.client.main -d -v localhost
-
-# Expected in journalctl output:
-#   hbc[<pid>]: hbc.main INFO: Starting hbc for <hostname> -> ['localhost']
-#   hbc[<pid>]: hbc.main INFO: hbc starting, sending heartbeat to localhost
-#   hbc[<pid>]: plugin.loader INFO: ...
-
-# Stop the daemon
-pkill -f "hbd.client.main"
-```
-
-- [ ] **Step 5: Commit**
-
-```bash
-git add hbd/client/main.py
-git commit -m "fix: reconfigure logging to syslog after daemonize() instead of no-op basicConfig"
-```

docs/superpowers/specs/2026-04-25-plugin-error-checking-design.md

@@ -1,92 +0,0 @@
-# Plugin Error Checking & Daemon Logging — Design Spec
-
-**Date:** 2026-04-25  
-**Scope:** hbc client — daemon mode logging, nagios_runner plugin robustness, PluginLoader messaging  
-**Files affected:** `hbd/client/main.py`, `hbd/client/plugins/nagios_runner.py`, `hbd/client/plugin.py`
-
----
-
-## 1. Daemon Mode Logging
-
-### Problem
-In `main()`, `logging.basicConfig()` is called before `daemonize()` (establishing a StreamHandler to stderr), then called again after `daemonize()`. The second call is a no-op — Python ignores `basicConfig()` when handlers are already configured. After daemonization, stderr is redirected to `/dev/null`, so all subsequent log output is silently discarded.
-
-The existing `syslog.openlog()` / `syslog.syslog()` calls (lines 666–668) write a single startup message but do not integrate with the `logging` system, so plugin and connection log messages never reach syslog.
-
-### Fix
-After `daemonize()`, explicitly reconfigure the root logger:
-
-1. Remove all existing handlers (they now write to `/dev/null`).
-2. Add `logging.handlers.SysLogHandler(address='/dev/log', facility=LOG_DAEMON)`.
-3. Set formatter: `hbc[%(process)d]: %(name)s %(levelname)s: %(message)s`
-4. Preserve the `log_level` already determined from `-v`/`-x` CLI flags.
-
-Remove the redundant `syslog.openlog()` / `syslog.syslog()` calls — the logging system handles routing.
-
-**Fallback:** If `/dev/log` does not exist (containers, some BSDs), fall back to `SysLogHandler(address=('localhost', 514))`. Log one warning (to stderr, before handlers are replaced) so the operator knows.
-
----
-
-## 2. Nagios Runner Improvements
-
-### 2a — Async Subprocess
-`_run_nagios_plugin()` is declared `async def` but calls `subprocess.run()` synchronously, blocking the event loop for the full command duration.
-
-**Fix:** Replace with `asyncio.create_subprocess_shell()` + `await proc.communicate()`. Enforce timeout with `asyncio.wait_for(..., timeout=self.timeout)` and catch `asyncio.TimeoutError`.
-
-### 2b — Stderr Capture
-Subprocess stderr is currently discarded (`capture_output=True` only captures stdout in the sync call; stderr content is lost).
-
-**Fix:** Pass `stderr=asyncio.subprocess.PIPE` to `create_subprocess_shell`. After `communicate()`, if stdout is empty but stderr has content, use stderr as the output message. If both have content, append stderr to the output for visibility.
-
-### 2c — Negative Return Codes
-A negative `returncode` means the process was killed by a signal (SIGKILL, OOM, etc.). The current code treats these as-is, which may produce unexpected status values.
-
-**Fix:** If `returncode < 0`, map to `NAGIOS_UNKNOWN` with message `"Process killed by signal {-returncode}"`.
-
-### 2d — Command Path Validation at Init
-`initialize()` currently only checks that the commands list is non-empty.
-
-**Fix:** For each command entry during `initialize()`:
-- Warn and skip the entry if `name` or `command` is missing.
-- Extract the executable (first whitespace-delimited token of the command string).
-- If the executable is an absolute path, check `os.path.isfile()` and `os.access(..., os.X_OK)`. Log a `WARNING` if either check fails.
-- Commands with relative paths or shell builtins are not checked (they may be on PATH) — just noted.
-- Validation warns only; all original entries in `self.commands` are retained and still attempted at collection time (where the existing missing-name/command guard already skips them). The plugin initializes successfully as long as the commands list is non-empty.
-
----
-
-## 3. PluginLoader Messaging
-
-### Problem
-When `initialize()` returns `False`, the loader always logs:
-> `WARNING: Plugin X failed initialization, skipping`
-
-This is alarming when the real reason is simply "no commands configured". There is no API to distinguish "not configured" from "genuinely broken".
-
-### Fix
-Add an optional `skip_reason` attribute to `Plugin.__init__()` (defaults to `None`).
-
-In `PluginLoader.load_from_directory()`, after `initialize()` returns `False`:
-- If `plugin.skip_reason` is set → `logger.info(f"Plugin {plugin.name} skipped: {plugin.skip_reason}")`
-- If `plugin.skip_reason` is `None` → `logger.warning(f"Plugin {plugin.name} failed initialization, skipping")` (existing behaviour)
-
-In `NagiosRunnerPlugin.initialize()`, when no commands are configured:
-```python
-self.skip_reason = "no commands configured (add nagios_runner.commands to config)"
-return False
-```
-
-Genuine failures (exceptions) continue to go through the existing `except` block in the loader, logging at `ERROR` with traceback — unchanged.
-
----
-
-## Decisions
-
-| Topic | Decision |
-|---|---|
-| Daemon log destination | syslog only (LOG_DAEMON facility) |
-| Syslog fallback | localhost:514 UDP if `/dev/log` absent |
-| Nagios result log level | INFO for all statuses (OK/WARNING/CRITICAL/UNKNOWN) |
-| Invalid command handling at init | Warn and continue; still attempt at collection time |
-| PluginLoader API change | `skip_reason` attribute on Plugin base class, checked by loader |

hbd/__init__.py

@@ -14,4 +14,4 @@ Install options:
 """
 
 __all__ = ["__version__"]
-__version__ = "5.1.9"
+__version__ = "5.4.0"

hbd/client/config.py

@@ -16,6 +16,12 @@ CLIENT_DEFAULTS = {
     "hb_port": 50003,          # Port where hbd servers listen
     "interval": 10,             # Heartbeat interval in seconds
 
+    # Host identity
+    "owner": None,             # Optional username to set as this host's owner on the server
+
+    # Security
+    "allow_remote_command": False,  # Execute shell commands received in CMD packets from the server
+
     # Runtime flags
     "foreground": False,
     "verbose": False,

hbd/client/main.py

@@ -21,6 +21,7 @@ from typing import Dict, List, Optional
 # Import protocol and config
 from .config import load_config
 from ..common.proto import dicttos, stodict
+from .. import __version__
 
 # Import plugin system
 from .plugin import PluginRegistry, PluginLoader, InfoPlugin, MonitorPlugin
@@ -36,6 +37,10 @@ dorestart = False
 shutdown_event: Optional[asyncio.Event] = None
 active_tasks: List[asyncio.Task] = []
 
+# Set from config in async_main. Off by default: a CMD packet is an unauthenticated
+# UDP datagram, so executing one must be opted into per host.
+allow_remote_command = False
+
 
 class AsyncConnection:
     """Async UDP connection to a heartbeat server."""
@@ -56,6 +61,9 @@ class AsyncConnection:
         self.transport: Optional[asyncio.DatagramTransport] = None
         self.protocol: Optional[asyncio.DatagramProtocol] = None
         self._dead = False
+        self._ever_opened = False
+        self._open_fail_count = 0   # consecutive failures before first success
+        self.request_info_event: asyncio.Event = asyncio.Event()
 
         self.logger = logging.getLogger(f"hbc.conn.{addr}")
 
@@ -73,6 +81,7 @@ class AsyncConnection:
                 lambda: HeartbeatProtocol(self),
                 family=self.af
             )
+            self._ever_opened = True
             self.logger.debug(f"Opened connection to {self.addr}:{self.port}")
             return True
         except Exception as e:
@@ -134,6 +143,9 @@ class AsyncConnection:
         
         self.ackcount += 1
         self.logger.debug(f"ACK received, RTT: {rtt:.1f}ms")
+        if msg.get("request_update"):
+            self.logger.info("server requested plugin info refresh")
+            self.request_info_event.set()
 
 
 class HeartbeatProtocol(asyncio.DatagramProtocol):
@@ -169,14 +181,13 @@ class HeartbeatProtocol(asyncio.DatagramProtocol):
             self.logger.error(f"Error processing datagram: {e}", exc_info=True)
     
     def error_received(self, exc):
-        """Handle protocol errors."""
+        """Handle protocol errors — close transport so the heartbeat sender retries."""
-        self.logger.warning(f"Protocol error on {self.connection.addr}: {exc} — dropping connection")
+        self.logger.warning(f"Protocol error on {self.connection.addr}: {exc} — will retry")
-        self.connection._dead = True
         self.connection.close()
 
 
 async def handle_command(conn: AsyncConnection, msg: dict):
-    """Execute a command received from server."""
+    """Execute a command received from server, if allow_remote_command is set."""
     import subprocess
 
     cmd = msg.get("cmd", "")
@@ -184,6 +195,15 @@ async def handle_command(conn: AsyncConnection, msg: dict):
         return
 
     logger = logging.getLogger("hbc.command")
+
+    if not allow_remote_command:
+        logger.warning(f"Refused command (allow_remote_command is false): {cmd}")
+        await conn.sendto({
+            "service": "command",
+            "msg": "Refused: allow_remote_command is false",
+        })
+        return
+
     logger.info(f"Executing command: {cmd}")
 
     try:
@@ -262,15 +282,51 @@ async def handle_update(conn: AsyncConnection, _msg: dict):  # pyright: ignore[r
 
 
 async def heartbeat_sender(conn: AsyncConnection, interval: int):
-    """Send periodic heartbeats.
+    """Send periodic heartbeats, retrying the connection if it is not open.
+
+    IPv6 connections that fail to open before their first successful send are
+    dropped after IPV6_EARLY_FAIL_LIMIT attempts so that a network without IPv6
+    does not keep a dead sender alive.  IPv4 connections are retried indefinitely.
 
     Args:
         conn: Connection to send on
         interval: Heartbeat interval in seconds
     """
     logger = logging.getLogger("hbc.heartbeat")
+    IPV6_EARLY_FAIL_LIMIT = 3
+
+    while running and not conn._dead:
+        # Ensure transport is open before attempting to send.
+        if not conn.transport:
+            opened = await conn.open()
+            if opened:
+                conn._open_fail_count = 0
+            else:
+                conn._open_fail_count += 1
+                # Drop an IPv6 connection that has never come up within the
+                # first few attempts — it is likely unavailable on this network.
+                if (not conn._ever_opened
+                        and conn.af == socket.AF_INET6
+                        and conn._open_fail_count >= IPV6_EARLY_FAIL_LIMIT):
+                    logger.warning(
+                        f"IPv6 connection to {conn.addr} unreachable after "
+                        f"{conn._open_fail_count} attempts, disabling"
+                    )
+                    conn._dead = True
+                    break
+                # Retry after the normal interval; IPv4 retries forever.
+                try:
+                    if shutdown_event:
+                        await asyncio.wait_for(shutdown_event.wait(), timeout=interval)
+                        break
+                    else:
+                        await asyncio.sleep(interval)
+                except asyncio.TimeoutError:
+                    pass
+                except asyncio.CancelledError:
+                    raise
+                continue
 
-    while running:
         try:
             msg = {
                 "acks": conn.ackcount,
@@ -279,19 +335,16 @@ async def heartbeat_sender(conn: AsyncConnection, interval: int):
             }
             await conn.sendto(msg, "HTB")
 
-        except Exception as e:
-            logger.error(f"Error sending heartbeat: {e}", exc_info=True)
         except asyncio.CancelledError:
             logger.debug("Heartbeat sender cancelled")
             raise
+        except Exception as e:
+            logger.error(f"Error sending heartbeat: {e}", exc_info=True)
 
         # Wait for next interval or shutdown event
         try:
             if shutdown_event:
-                await asyncio.wait_for(
+                await asyncio.wait_for(shutdown_event.wait(), timeout=interval)
-                    shutdown_event.wait(), 
-                    timeout=interval
-                )
                 break
             else:
                 await asyncio.sleep(interval)
@@ -302,6 +355,27 @@ async def heartbeat_sender(conn: AsyncConnection, interval: int):
             raise
 
 
+async def _info_plugin_refresh_loop(conn: AsyncConnection, info_plugins: List):
+    """Wait for server requests to re-send InfoPlugin data."""
+    logger = logging.getLogger("hbc.plugins")
+    while running:
+        await conn.request_info_event.wait()
+        if not running:
+            break
+        conn.request_info_event.clear()
+        logger.info("refreshing InfoPlugins on server request")
+        for plugin in info_plugins:
+            plugin._cache = None
+            try:
+                data = await plugin.collect()
+                if data:
+                    await conn.sendto(
+                        {"plugin": plugin.name, **data, "_interval": plugin.interval}, "PLG")
+                    logger.info(f"Resent {plugin.name} data")
+            except Exception as e:
+                logger.error(f"Error re-collecting {plugin.name}: {e}", exc_info=True)
+
+
 async def plugin_collector(conn: AsyncConnection, registry: PluginRegistry):
     """Collect and send plugin data.
 
@@ -317,8 +391,8 @@ async def plugin_collector(conn: AsyncConnection, registry: PluginRegistry):
         try:
             data = await plugin.collect()
             if data:
-                # Create PLG message with plugin name
+                # Create PLG message with plugin name and declared interval
-                plugin_msg = {"plugin": plugin.name, **data}
+                plugin_msg = {"plugin": plugin.name, **data, "_interval": plugin.interval}
                 await conn.sendto(plugin_msg, "PLG")
                 logger.info(f"Sent {plugin.name} data")
         except Exception as e:
@@ -333,24 +407,21 @@ async def plugin_collector(conn: AsyncConnection, registry: PluginRegistry):
     for plugin in monitor_plugins:
         by_interval[plugin.interval].append(plugin)
 
-    # Create tasks for each interval
+    # Create tasks for each interval; always include the info-refresh watcher
-    tasks = []
+    tasks = [asyncio.create_task(_info_plugin_refresh_loop(conn, info_plugins))]
     for interval, plugins in by_interval.items():
-        task = asyncio.create_task(
+        tasks.append(asyncio.create_task(
             plugin_collector_interval(conn, plugins, interval)
-        )
+        ))
-        tasks.append(task)
 
-    # Wait for all tasks
+    try:
-    if tasks:
+        await asyncio.gather(*tasks, return_exceptions=True)
-        try:
+    except asyncio.CancelledError:
-            await asyncio.gather(*tasks, return_exceptions=True)
+        logger.debug("Plugin collector cancelled, cancelling sub-tasks")
-        except asyncio.CancelledError:
+        for task in tasks:
-            logger.debug("Plugin collector cancelled, cancelling sub-tasks")
+            if not task.done():
-            for task in tasks:
+                task.cancel()
-                if not task.done():
+        raise
-                    task.cancel()
-            raise
 
 
 async def plugin_collector_interval(
@@ -373,7 +444,7 @@ async def plugin_collector_interval(
                 data = await plugin.collect()
                 if data:
                     # Don't use encode_plugin_data - create dict directly
-                    plugin_msg = {"plugin": plugin.name, **data}
+                    plugin_msg = {"plugin": plugin.name, **data, "_interval": plugin.interval}
                     await conn.sendto(plugin_msg, "PLG")
                     logger.debug(f"Sent {plugin.name} data")
             except asyncio.CancelledError:
@@ -427,16 +498,14 @@ async def cleanup(connections: List[AsyncConnection]):
     logger = logging.getLogger("hbc.cleanup")
     logger.info("Cleaning up connections")
     
-    for conn in connections:
+    target = next((c for c in connections if c.transport), connections[0] if connections else None)
+    # A SIGHUP restart is not a host shutdown, so don't announce one.
+    if target and send_shutdown and not dorestart:
         try:
-            msg = {
+            await target.sendto({"shutdown": 1, "acks": target.ackcount})
-                "shutdown": 1,
-                "acks": conn.ackcount
-            }
-            await conn.sendto(msg)
         except Exception as e:
             logger.error(f"Error sending shutdown: {e}")
-        
+    for conn in connections:
         conn.close()
     
     # Give messages time to send
@@ -445,11 +514,12 @@ async def cleanup(connections: List[AsyncConnection]):
 
 async def async_main(args, config):
     """Async main function."""
-    global running, shutdown_event, active_tasks
+    global running, shutdown_event, active_tasks, send_shutdown, allow_remote_command
 
     # Create shutdown event
     shutdown_event = asyncio.Event()
     active_tasks = []
+    allow_remote_command = bool(config.get("allow_remote_command", False))
     
     logger = logging.getLogger("hbc.main")
     
@@ -462,47 +532,63 @@ async def async_main(args, config):
     hb_port = config.get("hb_port", PORT)
     interval = config.get("interval", INTERVAL)
     
-    logger.info(f"Starting hbc for {iam} -> {hb_hosts}")
+    logger.info(f"hbc {__version__} on {iam} -> {hb_hosts} port={hb_port}, interval={interval}s")
-    logger.info(f"Port: {hb_port}, Interval: {interval}s")
+    if allow_remote_command:
+        logger.warning("allow_remote_command is true — CMD packets from the server will be executed")
+    
+    af_filter = (socket.AF_INET if getattr(args, "ipv4_only", False)
+                 else socket.AF_INET6 if getattr(args, "ipv6_only", False)
+                 else 0)
 
     # Create connections
     connections = []
     conn_id = 1
+    _retry_delay = 5
 
-    for host in hb_hosts:
+    while running and not connections:
-        try:
+        for host in hb_hosts:
-            addrs = socket.getaddrinfo(host, hb_port, 0, 0, socket.SOL_UDP)
+            try:
-        except socket.gaierror as e:
+                addrs = socket.getaddrinfo(host, hb_port, af_filter, 0, socket.SOL_UDP)
-            logger.error(f"Cannot resolve {host}: {e}")
+            except socket.gaierror as e:
-            continue
+                logger.warning(f"Cannot resolve {host}: {e} — retrying in {_retry_delay}s")
-        
+                continue
-        for addr_info in addrs:
+            for addr_info in addrs:
-            af = addr_info[0]
+                af = addr_info[0]
-            addr = addr_info[4][0]
+                addr = addr_info[4][0]
-            
+                conn = AsyncConnection(conn_id, addr, hb_port, af, iam)
-            conn = AsyncConnection(conn_id, addr, hb_port, af, iam)
+                if not await conn.open():
-            if await conn.open():
+                    logger.warning(f"Initial open to {addr} failed, heartbeat sender will retry")
                 connections.append(conn)
                 conn_id += 1
+        if not connections:
+            try:
+                if shutdown_event:
+                    await asyncio.wait_for(shutdown_event.wait(), timeout=_retry_delay)
+                else:
+                    await asyncio.sleep(_retry_delay)
+            except asyncio.TimeoutError:
+                pass
+            _retry_delay = min(_retry_delay * 2, 60)
 
     if not connections:
-        logger.error("No connections established")
         return 1
 
     logger.info(f"Created {len(connections)} connections")
     
     # Send boot/message if requested
+    send_shutdown = False
     if args.boot or args.message:
         boot_msg = {}
         if args.boot:
             boot_msg["boot"] = 1
+            send_shutdown = True
         if args.message:
             boot_msg["service"] = "service"
             boot_msg["msg"] = args.message
         
         boot_msg["acks"] = 0
-        for conn in connections:
+        target = next((c for c in connections if c.transport), connections[0])
-            await conn.sendto(boot_msg)
+        await target.sendto(boot_msg)
         
         if args.message and not args.daemon:
             # Message-only mode
@@ -525,6 +611,13 @@ async def async_main(args, config):
     for sig in (signal.SIGTERM, signal.SIGINT):
         loop.add_signal_handler(sig, stop)
 
+    def _sighup():
+        global dorestart
+        dorestart = True
+        stop()
+
+    loop.add_signal_handler(signal.SIGHUP, _sighup)
+    
     # Start async tasks
     # Heartbeat senders (one per connection)
     for conn in connections:
@@ -662,6 +755,9 @@ def build_parser():
         default=0,
         help="Increase debug level"
     )
+    af_group = parser.add_mutually_exclusive_group()
+    af_group.add_argument("-4", dest="ipv4_only", action="store_true", help="Use IPv4 only")
+    af_group.add_argument("-6", dest="ipv6_only", action="store_true", help="Use IPv6 only")
     parser.add_argument(
         "hosts",
         nargs="+",
@@ -695,7 +791,7 @@ def main(argv=None):
     
     # Daemonize if requested
     if args.daemon:
-        print("Daemonizing...")
+        logging.info("Daemonizing...")
         daemonize()
         _reconfigure_logging_for_daemon(log_level)
         logging.info(f"hbc starting, sending heartbeat to {', '.join(args.hosts)}")
@@ -713,7 +809,10 @@ def main(argv=None):
     # Handle restart
     if dorestart:
         logging.info("Restarting...")
-        os.execv(sys.argv[0], sys.argv)
+        # Drop -b/--boot so the re-exec'd process doesn't re-announce a boot;
+        # a SIGHUP restart is not a host reboot.
+        restart_argv = [sys.argv[0]] + [a for a in sys.argv[1:] if a not in ("-b", "--boot")]
+        os.execv(restart_argv[0], restart_argv)
     
     sys.exit(exit_code)
 

hbd/client/plugin.py

@@ -364,7 +364,10 @@ class PluginLoader:
                     
                     # Instantiate plugin with config — check plugins subdict first,
                     # then top-level keys (e.g. nagios_runner: ... at root of config).
-                    plugin_instance_config = plugins_subconfig.get(obj.name) or raw_config.get(obj.name, {})
+                    plugin_instance_config = dict(plugins_subconfig.get(obj.name) or raw_config.get(obj.name) or {})
+                    # Propagate top-level owner so os_info (and any future plugin) can report it.
+                    if "owner" in raw_config and "owner" not in plugin_instance_config:
+                        plugin_instance_config["owner"] = raw_config["owner"]
                     plugin = obj(config=plugin_instance_config)
                     
                     # Initialize plugin

hbd/client/plugins/cpu_monitor.py

@@ -119,6 +119,13 @@ class CPUMonitorPlugin(MonitorPlugin):
             except Exception as e:
                 self.logger.debug(f"Could not get CPU times: {e}")
 
+            # Uptime in seconds
+            try:
+                import time
+                data["uptime_seconds"] = int(time.time() - self.psutil.boot_time())
+            except Exception as e:
+                self.logger.debug(f"Could not get uptime: {e}")
+            
             self.logger.debug(
                 f"Collected CPU metrics: {data.get('cpu_percent', 'N/A')}% usage"
             )

hbd/client/plugins/filesystem_info.py

@@ -127,7 +127,7 @@ class FilesystemInfoPlugin(InfoPlugin):
                     try:
                         # Maximum filename length
                         max_name = os.pathconf(partition.mountpoint, 'PC_NAME_MAX')
-                        if max_name:
+                        if max_name is not None:
                             fs_info['maxfile'] = max_name
                     except (OSError, ValueError):
                         pass
@@ -135,7 +135,7 @@ class FilesystemInfoPlugin(InfoPlugin):
                     try:
                         # Maximum path length
                         max_path = os.pathconf(partition.mountpoint, 'PC_PATH_MAX')
-                        if max_path:
+                        if max_path is not None:
                             fs_info['maxpath'] = max_path
                     except (OSError, ValueError):
                         pass

hbd/client/plugins/memory_monitor.py

@@ -14,6 +14,24 @@ except ImportError:
 
 from hbd.client.plugin import MonitorPlugin
 
+
+def _zfs_arc_bytes() -> int:
+    """Return current ZFS ARC size in bytes, or 0 if ZFS is not present.
+
+    ZFS ARC is reclaimable but is not included in MemAvailable by the Linux
+    kernel (it is not in SReclaimable), so it would otherwise be counted as
+    used memory.
+    """
+    try:
+        with open("/proc/spl/kstat/zfs/arcstats") as fh:
+            for line in fh:
+                parts = line.split()
+                if len(parts) >= 3 and parts[0] == "size":
+                    return int(parts[2])
+    except (OSError, ValueError):
+        pass
+    return 0
+
 logger = logging.getLogger(__name__)
 
 
@@ -101,11 +119,21 @@ class MemoryMonitorPlugin(MonitorPlugin):
         
         # Virtual (physical) memory statistics
         vmem = psutil.virtual_memory()
+
+        # psutil's available already excludes page cache / file buffers
+        # (uses MemAvailable on Linux). Add ZFS ARC on top because the kernel
+        # does not include it in SReclaimable / MemAvailable even though it is
+        # reclaimable.
+        arc_bytes = _zfs_arc_bytes()
+        available = min(vmem.available + arc_bytes, vmem.total)
+        used = vmem.total - available
+        percent = round(used / vmem.total * 100, 1) if vmem.total else 0.0
+
         metrics['memory_total'] = vmem.total
-        metrics['memory_available'] = vmem.available
+        metrics['memory_available'] = available
-        metrics['memory_used'] = vmem.used
+        metrics['memory_used'] = used
         metrics['memory_free'] = vmem.free
-        metrics['memory_percent'] = vmem.percent
+        metrics['memory_percent'] = percent
         
         # Platform-specific memory details
         if hasattr(vmem, 'active'):

hbd/client/plugins/nagios_runner.py

@@ -31,16 +31,13 @@ from hbd.client.plugin import MonitorPlugin
 
 
 # Nagios exit codes
-NAGIOS_OK = 0
-NAGIOS_WARNING = 1
-NAGIOS_CRITICAL = 2
 NAGIOS_UNKNOWN = 3
 
 STATUS_NAMES = {
-    NAGIOS_OK: "OK",
+    0: "OK",
-    NAGIOS_WARNING: "WARNING",
+    1: "WARNING",
-    NAGIOS_CRITICAL: "CRITICAL",
+    2: "CRITICAL",
-    NAGIOS_UNKNOWN: "UNKNOWN"
+    3: "UNKNOWN",
 }
 
 
@@ -129,9 +126,6 @@ class NagiosRunnerPlugin(MonitorPlugin):
         """
         results = {}
 
-        # Track overall status (worst status wins)
-        worst_status = NAGIOS_OK
-        
         for cmd_config in self.commands:
             name = cmd_config.get("name")
             command = cmd_config.get("command")
@@ -149,10 +143,6 @@ class NagiosRunnerPlugin(MonitorPlugin):
                 results[f"{name}_status_code"] = status_code
                 results[f"{name}_output"] = output
 
-                # Track worst status
-                if status_code > worst_status:
-                    worst_status = status_code
-                
                 # Parse and add performance data
                 if perfdata:
                     for metric_name, metric_value in perfdata.items():
@@ -167,12 +157,6 @@ class NagiosRunnerPlugin(MonitorPlugin):
                 results[f"{name}_status"] = "ERROR"
                 results[f"{name}_status_code"] = NAGIOS_UNKNOWN
                 results[f"{name}_output"] = str(e)
-                worst_status = NAGIOS_UNKNOWN
-        
-        # Add overall status
-        results["overall_status"] = STATUS_NAMES.get(worst_status, "UNKNOWN")
-        results["overall_status_code"] = worst_status
-        results["plugin_count"] = len(self.commands)
 
         return results
     

hbd/client/plugins/os_info.py

@@ -62,6 +62,9 @@ class OSInfoPlugin(InfoPlugin):
                 "hbc_version": hbc_version,
                 "hbc_type": "full",
             }
+            if self.config.get("owner"):
+                self.logger.debug(f"Adding owner from config: {self.config['owner']}")
+                data["owner"] = self.config["owner"]
             
             # Add Linux-specific distribution info
             if platform.system() == "Linux":

hbd/client/plugins/ping_monitor.py

@@ -13,12 +13,8 @@ plugins:
     count: 3              # ICMP packets per ping run (default 3)
     timeout: 5            # seconds before a host is considered unreachable (default 5)
     hosts:
-      8.8.8.8:
+      - 8.8.8.8
-        warning: 20.0     # ms
+      - 192.168.1.1
-        critical: 100.0   # ms
-      192.168.1.1:
-        warning: 5.0
-        critical: 20.0
 ```
 
 Reported metrics per host (metric key uses the hostname with dots/colons replaced

hbd/client/plugins/zfs_monitor.py

@@ -0,0 +1,140 @@
+"""
+ZFS pool monitoring plugin for Heartbeat.
+
+Collects per-pool health, capacity, and cumulative I/O statistics via zpool(8).
+"""
+
+import asyncio
+import logging
+import shutil
+from typing import Any, Dict, List, Optional
+
+from hbd.client.plugin import MonitorPlugin
+
+logger = logging.getLogger(__name__)
+
+
+def _int(s: str) -> Optional[int]:
+    try:
+        return int(s.strip().rstrip("KMGTkBkmgt%x"))
+    except (ValueError, AttributeError):
+        return None
+
+
+def _float(s: str) -> Optional[float]:
+    try:
+        return float(s.strip().rstrip("%x"))
+    except (ValueError, AttributeError):
+        return None
+
+
+class ZFSMonitorPlugin(MonitorPlugin):
+    """Monitor ZFS pool health, capacity, and I/O statistics.
+
+    Collects per pool:
+    - health: ONLINE, DEGRADED, FAULTED, etc.
+    - size / alloc / free: total, allocated and free bytes
+    - capacity: percentage used (0-100)
+    - frag: fragmentation percentage
+    - dedup: deduplication ratio
+    - read_ops / write_ops: cumulative I/O operations since last boot/clear
+    - read_bw / write_bw: cumulative bytes transferred since last boot/clear
+
+    Configuration:
+        interval: collection interval in seconds (default: 300)
+        pools: list of pool names to monitor (default: all)
+    """
+
+    name = "zfs_monitor"
+    description = "ZFS pool health, capacity, and I/O statistics"
+    interval = 300
+
+    def __init__(self, config: Optional[Dict[str, Any]] = None):
+        super().__init__(config)
+        self.interval = self.config.get("interval", 300)
+        self._pools_filter: Optional[List[str]] = self.config.get("pools", None)
+
+    async def initialize(self) -> bool:
+        if not shutil.which("zpool"):
+            self.skip_reason = "zpool not found"
+            return False
+        logger.info("ZFS monitor initialized (interval: %ds)", self.interval)
+        return True
+
+    async def _run(self, *args: str) -> List[str]:
+        """Run a command and return its stdout lines, or [] on error."""
+        try:
+            proc = await asyncio.create_subprocess_exec(
+                *args,
+                stdout=asyncio.subprocess.PIPE,
+                stderr=asyncio.subprocess.DEVNULL,
+            )
+            stdout, _ = await asyncio.wait_for(proc.communicate(), timeout=15)
+            return stdout.decode(errors="replace").splitlines()
+        except (FileNotFoundError, asyncio.TimeoutError) as exc:
+            logger.warning("zfs_monitor: %s: %s", args[0], exc)
+            return []
+
+    async def _zpool_list(self) -> Dict[str, Dict]:
+        """Return per-pool health and capacity from `zpool list`."""
+        lines = await self._run(
+            "zpool", "list", "-H", "-p",
+            "-o", "name,health,size,alloc,free,cap,frag,dedup",
+        )
+        pools: Dict[str, Dict] = {}
+        for line in lines:
+            parts = line.split("\t")
+            if len(parts) < 8:
+                continue
+            name = parts[0].strip()
+            if self._pools_filter and name not in self._pools_filter:
+                continue
+            health = parts[1].strip()
+            if health == "ONLINE":
+                status = 0
+            elif health in ("DEGRADED", "ONLINE with errors"):
+                status = 1
+            elif health in ("FAULTED", "OFFLINE", "UNAVAIL"):
+                status = 2
+            else:
+                status = 3  # unknown status
+            pools[name] = {
+                "health":    health,
+                "status": status,
+                "size":      _int(parts[2]),
+                "alloc":     _int(parts[3]),
+                "free":      _int(parts[4]),
+                "capacity":  _float(parts[5]),
+                "frag":      _float(parts[6]),
+                "dedup":     _float(parts[7]),
+            }
+        return pools
+
+    async def _zpool_iostat(self) -> Dict[str, Dict]:
+        """Return per-pool cumulative I/O counters from `zpool iostat`."""
+        lines = await self._run("zpool", "iostat", "-H", "-p")
+        io: Dict[str, Dict] = {}
+        for line in lines:
+            parts = line.split("\t")
+            if len(parts) < 7:
+                continue
+            name = parts[0].strip()
+            if not name or name.startswith(" "):
+                continue
+            io[name] = {
+                "read_ops": _int(parts[3]),
+                "write_ops": _int(parts[4]),
+                "read_bw":  _int(parts[5]),
+                "write_bw": _int(parts[6]),
+            }
+        return io
+
+    async def _collect_metrics(self) -> Dict[str, Any]:
+        pools, io = await asyncio.gather(self._zpool_list(), self._zpool_iostat())
+        for name, stats in io.items():
+            if name in pools:
+                pools[name].update(stats)
+        return {"pools": pools}
+
+
+plugin = ZFSMonitorPlugin

hbd/config_thresholds_example.yaml

@@ -134,6 +134,31 @@ thresholds:
           hysteresis: 0.1
           enabled: true
   
+  # ----------------------------------------------------------------------------
+  # ZFS Monitor Thresholds
+  # ----------------------------------------------------------------------------
+  zfs_monitor:
+    # Pool health check — built-in default; shown here for reference/override.
+    # status is 0 (ONLINE) or 1 (DEGRADED) or 2 (SUSPENDED, FAULTED, UNAVAIL…).
+    # Use '*' to apply the same rule to every pool, or name a specific pool.
+    pools:
+      '*':
+        status:
+          warning: 1           # Alert WARNING when pool is DEGRADED
+          critical: 2           # Alert CRITICAL when pool is SUSPENDED/FAULTED/UNAVAIL
+          operator: ">="
+          hysteresis: 0.0       # No hysteresis — a degraded pool is always alerting
+          grace: 0              # Fire immediately — don't wait for a second collection
+          display: "ZFS pool {pool_name} is {health}"
+
+      # Per-pool capacity thresholds (optional; add pools you care about)
+      # tank:
+      #   capacity:
+      #     warning: 75.0       # Warn at 75% used
+      #     critical: 90.0      # Critical at 90% used
+      #     operator: ">"
+      #     hysteresis: 0.05
+
   # ----------------------------------------------------------------------------
   # Network Monitor Thresholds
   # ----------------------------------------------------------------------------

hbd/server/config.py

@@ -27,21 +27,26 @@ SERVER_DEFAULTS = {
 
     # Monitoring settings
     "interval": 20,             # Expected heartbeat interval (for server checks)
-    "grace": 2,                 # Grace multiplier (interval * grace = timeout)
+    "grace": 2,                 # Grace period (extra seconds before notifying after a missed heartbeat)
     "threshold_renotify_interval": 3600,  # Seconds between threshold re-notifications
 
+    # Flap detection (0 in either key disables it)
+    "flap_count": 5,            # Warning/critical notifications within flap_interval that trip flapping
+    "flap_interval": 10,        # Minutes: the counting window, and the quiet window after an OK
+
     # User management
     "users": {},                # username -> {full_name, avatar, password, admin, notification_channels}
     "default_owner": None,      # Username that owns hosts with no explicit owner
 
+    # OAuth2 providers
+    "oauth": {},                 # oauth.gitea.{url,client_id,client_secret}
+
     # Host management
     "hosts": {},                # Unified host definitions
-    "dyndnshosts": [],          # Hosts with dynamic DNS (legacy)
+    "dyndomains": ["example.org"], # Domains to update via nsupdate when a host with dyndns: true is updated
-    "drophosts": [],            # Hosts to ignore
-    "dyndomains": ["wrede.org"],
 
     # DNS updates
-    "nsupdate_bin": "/usr/bin/nsupdate",
+    "nsupdate_bin": "/usr/bin/nsupdate", # Path to nsupdate binary
     
     # WebSocket settings
     "ws_port": 50005, 
@@ -76,9 +81,13 @@ THRESHOLD_DEFAULTS = {
                 }
             },
             'memory_monitor': {
-                'percent': {
+                'memory_percent': {
                     'warning': 85.0,
                     'critical': 95.0
+                },
+                'swap_percent': {
+                    'warning': 40.0,
+                    'critical': 75.0
                 }
             },
             'disk_monitor': {
@@ -95,7 +104,31 @@ THRESHOLD_DEFAULTS = {
                 'warning': 200,
                 'critical': 250.0,
                 'count': 3  # Optional: number of consecutive breaches before alerting
-            }
+            },
+            'nagios_runner': {
+                'status_code': {
+                    'display': '{check_name} {output}',
+                    'operator': "nagios"
+                }
+            },
+            'zfs_monitor': {
+                'pools': {
+                    '*': {
+                        'status': {
+                            'warning': 1,
+                            'critical': 2,
+                            'operator': '>=',
+                            'hysteresis': 0.0,
+                            'grace': 0,
+                            'display': 'ZFS pool {pool_name} is {health}'
+                        },
+                        'capacity': {
+                            'warning': 80.0,
+                            'critical': 90.0,
+                        }
+                    }
+                }
+            },
         }
     }
 
@@ -219,43 +252,26 @@ def get_watchhosts(config):
     """Extract watched hostnames from config (hosts with watch: true).
 
     Returns:
-        List of hostnames to watch
+#        List of hostnames to watch
     """
     watchhosts = []
     hosts_config = config.get("hosts", {})
     if isinstance(hosts_config, dict):
         for host_name, host_attrs in hosts_config.items():
-            if isinstance(host_attrs, dict) and host_attrs.get("watch", False):
+            if isinstance(host_attrs, dict) and host_attrs.get("watch", True):
                 watchhosts.append(host_name)
     return watchhosts
 
 
 def get_dyndnshosts(config):
-    """Extract dyndnshosts from config, supporting both new and legacy formats.
+    """Return hostnames that have a dyndns setting in the hosts section."""
-    
+    hosts_config = config.get("hosts", {})
-    Args:
+    if not isinstance(hosts_config, dict):
-        config: Configuration dictionary
+        return []
-        
+    return [
-    Returns:
+        name for name, attrs in hosts_config.items()
-        List of hostnames with dynamic DNS
+        if isinstance(attrs, dict) and attrs.get("dyndns")
-    """
+    ]
-    dyndnshosts = []
-    
-    # New format: hosts section with dyndns attribute
-    if "hosts" in config:
-        hosts_config = config["hosts"]
-        if isinstance(hosts_config, dict):
-            for host_name, host_attrs in hosts_config.items():
-                if isinstance(host_attrs, dict) and host_attrs.get("dyndns", False):
-                    dyndnshosts.append(host_name)
-    
-    # Legacy format: dyndnshosts list/set
-    if "dyndnshosts" in config:
-        legacy_dyndnshosts = config.get("dyndnshosts", [])
-        if isinstance(legacy_dyndnshosts, (list, set)):
-            dyndnshosts.extend(legacy_dyndnshosts)
-    
-    return list(set(dyndnshosts))  # Remove duplicates
 
 
 def get_host_config(config, hostname):
@@ -303,7 +319,7 @@ def get_host_access(config, hostname) -> dict:
     """
     host_cfg = get_host_config(config, hostname)
 
-    owner = host_cfg.get("owner") or get_default_owner(config)
+    owner = host_cfg.get("owner") # or get_default_owner(config)
 
     managers = host_cfg.get("managers", [])
     if isinstance(managers, str):

hbd/server/config_access.py

@@ -0,0 +1,168 @@
+"""Ownership rules for config-file entities (hosts, channels, threshold configs).
+
+Rule: an entry whose config dict has a non-empty ``owner`` is private to that
+owner — visible, usable, and editable only by the owner (and admins).  An
+entry with no ``owner`` is global: usable by everyone, editable by admins only.
+
+The scoped-merge helpers implement the non-admin save path for
+``POST /api/0/config``: the caller's visible subset of a section is replaced
+by the submitted payload; everything else is preserved untouched.
+"""
+
+from typing import Any, Dict
+
+
+class ScopedMergeError(Exception):
+    """A non-admin payload violated an ownership rule; message names the entry."""
+
+
+def is_global(cfg: Any) -> bool:
+    """True when *cfg* has no owner (usable by everyone)."""
+    return not (isinstance(cfg, dict) and cfg.get("owner"))
+
+
+def user_can_use(cfg: Any, username: str) -> bool:
+    """True when *username* may use/see this entry (global or own)."""
+    return is_global(cfg) or cfg.get("owner") == username
+
+
+def _as_list(value: Any) -> list:
+    if value is None:
+        return []
+    if isinstance(value, str):
+        return [value]
+    return list(value)
+
+
+def user_hosts(hosts_cfg: Any, username: str) -> Dict[str, Any]:
+    """Subset of *hosts_cfg* where *username* is owner or manager."""
+    result = {}
+    for name, cfg in (hosts_cfg or {}).items():
+        if not isinstance(cfg, dict):
+            continue
+        if cfg.get("owner") == username or username in _as_list(cfg.get("managers")):
+            result[name] = cfg
+    return result
+
+
+def user_channels(channels_cfg: Any, username: str) -> Dict[str, Any]:
+    """Subset of channels usable by *username*: global + own."""
+    return {
+        name: cfg for name, cfg in (channels_cfg or {}).items()
+        if isinstance(cfg, dict) and user_can_use(cfg, username)
+    }
+
+
+def user_threshold_configs(threshold_cfgs: Any, username: str) -> Dict[str, Any]:
+    """Subset of threshold configs usable by *username*: global + own."""
+    return {
+        name: cfg for name, cfg in (threshold_cfgs or {}).items()
+        if isinstance(cfg, dict) and user_can_use(cfg, username)
+    }
+
+
+_HOST_SETTING_KEYS = ("watch", "dyndns", "notification_channels", "threshold_config")
+_HOST_ACCESS_KEYS = ("owner", "managers", "monitors")
+_HOST_ALLOWED_KEYS = frozenset(_HOST_SETTING_KEYS) | frozenset(_HOST_ACCESS_KEYS)
+
+
+def _check_added_assignments(host_name: str, old_cfg: Any, new_cfg: dict,
+                             key: str, usable: set) -> None:
+    """Every value under *key* not already on the host must be in *usable*."""
+    old_vals = set(_as_list((old_cfg or {}).get(key)))
+    for v in _as_list(new_cfg.get(key)):
+        if v not in old_vals and v not in usable:
+            raise ScopedMergeError(
+                f"host {host_name!r}: {key} entry {v!r} is not available to you")
+
+
+def merge_hosts_scoped(existing: Any, payload: Any, username: str,
+                       channels_cfg: Any, threshold_cfgs: Any) -> Dict[str, Any]:
+    """Return a new hosts section with *username*'s visible subset replaced by *payload*.
+
+    Hosts the user cannot see are preserved untouched.  Visible hosts missing
+    from the payload are deleted (owners only).  Managers may change setting
+    keys but not access keys and may not delete.  New hosts get their owner
+    forced to *username*.  Newly added channel/threshold assignments must be
+    global or owned by the user.  Violations raise ScopedMergeError.
+    """
+    existing = existing or {}
+    payload = payload or {}
+    visible = user_hosts(existing, username)
+    usable_channels = set(user_channels(channels_cfg, username))
+    usable_tcs = set(user_threshold_configs(threshold_cfgs, username)) | {"default"}
+
+    result: Dict[str, Any] = {n: c for n, c in existing.items() if n not in visible}
+
+    for name, entry in payload.items():
+        if not isinstance(entry, dict):
+            raise ScopedMergeError(f"host {name!r}: invalid entry")
+        unknown = set(entry) - _HOST_ALLOWED_KEYS
+        if unknown:
+            raise ScopedMergeError(
+                f"host {name!r}: fields not permitted: {', '.join(sorted(unknown))}")
+        if name in existing and name not in visible:
+            raise ScopedMergeError(f"host {name!r}: you are not an owner or manager")
+
+        old = visible.get(name)
+        if old is None:
+            new_cfg = dict(entry)
+            new_cfg["owner"] = username
+        else:
+            new_cfg = dict(old)
+            for key in _HOST_SETTING_KEYS:
+                if key in entry:
+                    new_cfg[key] = entry[key]
+                else:
+                    new_cfg.pop(key, None)
+            if old.get("owner") == username:
+                for key in _HOST_ACCESS_KEYS:
+                    if key in entry:
+                        new_cfg[key] = entry[key]
+                    else:
+                        new_cfg.pop(key, None)
+            else:
+                if "owner" in entry and (entry.get("owner") or None) != (old.get("owner") or None):
+                    raise ScopedMergeError(
+                        f"host {name!r}: only the owner may change ownership")
+                for key in ("managers", "monitors"):
+                    if key in entry and set(_as_list(entry[key])) != set(_as_list(old.get(key))):
+                        raise ScopedMergeError(
+                            f"host {name!r}: only the owner may change {key}")
+
+        _check_added_assignments(name, old, new_cfg, "notification_channels", usable_channels)
+        _check_added_assignments(name, old, new_cfg, "threshold_config", usable_tcs)
+        result[name] = new_cfg
+
+    for name, old in visible.items():
+        if name not in payload and old.get("owner") != username:
+            raise ScopedMergeError(f"host {name!r}: only the owner may delete a host")
+
+    return result
+
+
+def merge_threshold_configs_scoped(existing: Any, payload: Any,
+                                   username: str) -> Dict[str, Any]:
+    """Return a new threshold_configs section with the user's own configs
+    replaced by *payload*.
+
+    Global and foreign-owned configs are preserved and may not appear in the
+    payload ('default' included).  Own configs missing from the payload are
+    deleted.  Every payload entry gets its owner forced to *username*.
+    """
+    existing = existing or {}
+    payload = payload or {}
+    result: Dict[str, Any] = {
+        n: c for n, c in existing.items()
+        if not (isinstance(c, dict) and c.get("owner") == username)
+    }
+    for name, entry in payload.items():
+        if name == "default":
+            raise ScopedMergeError("threshold config 'default' is global and admin-managed")
+        old = existing.get(name)
+        if old is not None and (not isinstance(old, dict) or old.get("owner") != username):
+            raise ScopedMergeError(f"threshold config {name!r}: not owned by you")
+        new_cfg = dict(entry) if isinstance(entry, dict) else {}
+        new_cfg["owner"] = username
+        result[name] = new_cfg
+    return result

hbd/server/configio.py

@@ -0,0 +1,137 @@
+"""YAML round-trip read/write for .hb.yaml, with backup and atomic writes."""
+
+import glob
+import os
+import threading
+from datetime import datetime
+
+from ruamel.yaml import YAML
+
+_write_lock = threading.Lock()
+
+
+def _make_yaml() -> YAML:
+    y = YAML()
+    y.preserve_quotes = True
+    return y
+
+# Top-level keys managed by the 'server' logical section
+_SERVER_KEYS = [
+    "hbd_port", "hbd_host", "ws_port", "wss_port", "hb_port",
+    "interval", "grace", "base_url", "threshold_renotify_interval",
+    "flap_count", "flap_interval",
+    "logfile", "pidfile", "pickfile", "journal_enabled", "journal_dir",
+    "journal_max_size", "journal_max_backups", "default_owner",
+    "default_threshold_config",
+]
+
+# Top-level keys managed by the 'dns' logical section
+_DNS_KEYS = ["nsupdate_bin", "rndc_key", "dyndomains"]
+
+
+def read_roundtrip(path: str):
+    """Load .hb.yaml with ruamel.yaml, preserving comments and ordering."""
+    with open(path, "r", encoding="utf-8") as f:
+        return _make_yaml().load(f)
+
+
+def write_config(path: str, data) -> None:
+    """Backup current file then atomically write data.
+
+    Backup naming: {path}.bak.YYYYMMDD-HHMMSS
+    Rotation: keep the 10 most recent backups, delete older ones.
+    Atomic write: write to {path}.tmp then os.replace({path}.tmp, path).
+    Acquires _write_lock for the full backup+write sequence.
+    """
+    with _write_lock:
+        ts = datetime.now().strftime("%Y%m%d-%H%M%S")
+        backup_path = f"{path}.bak.{ts}"
+        n = 0
+        while os.path.exists(backup_path):
+            n += 1
+            backup_path = f"{path}.bak.{ts}-{n}"
+        orig_mode = None
+        if os.path.exists(path):
+            orig_mode = os.stat(path).st_mode
+            with open(path, "rb") as src, open(backup_path, "wb") as dst:
+                dst.write(src.read())
+            os.chmod(backup_path, orig_mode)
+        backups = sorted(glob.glob(f"{path}.bak.*"), reverse=True)
+        for old in backups[10:]:
+            os.unlink(old)
+        tmp = f"{path}.tmp"
+        try:
+            with open(tmp, "w", encoding="utf-8") as f:
+                _make_yaml().dump(data, f)
+            if orig_mode is not None:
+                os.chmod(tmp, orig_mode)
+            os.replace(tmp, path)
+        except Exception:
+            try:
+                os.unlink(tmp)
+            except OSError:
+                pass
+            raise
+
+
+def list_backups(path: str) -> list:
+    """Return backup paths sorted newest-first."""
+    return sorted(glob.glob(f"{path}.bak.*"), reverse=True)
+
+
+def apply_structured_section(data, section: str, values: dict) -> None:
+    """Merge a dict of scalar/list values into data for the named logical section.
+
+    For 'server': updates each known key individually, preserving comments on
+    unchanged keys. For 'users': replaces the entire users dict.
+    """
+    if section == "server":
+        for key in _SERVER_KEYS:
+            if key in values:
+                data[key] = values[key]
+    elif section == "dns":
+        for key in _DNS_KEYS:
+            if key in values:
+                data[key] = values[key]
+            else:
+                data.pop(key, None)
+    elif section == "users":
+        data["users"] = values
+    elif section == "hosts":
+        data["hosts"] = values
+    else:
+        raise ValueError(f"Unknown structured section: {section!r}")
+
+
+def apply_channel(data, name: str, channel_cfg: dict) -> None:
+    """Insert or replace a single notification channel entry, preserving others."""
+    if not data.get("notification_channels"):
+        data["notification_channels"] = {}
+    data["notification_channels"][name] = channel_cfg
+
+
+def delete_channel(data, name: str) -> None:
+    """Remove a notification channel by name. No-op if not found."""
+    nc = data.get("notification_channels") or {}
+    nc.pop(name, None)
+
+
+def apply_yaml_section(data, section: str, yaml_text: str) -> None:
+    """Replace the named logical section by parsing yaml_text."""
+    parsed = _make_yaml().load(yaml_text)
+    if section == "notification_channels":
+        data["notification_channels"] = parsed
+    elif section == "thresholds":
+        data["threshold_configs"] = parsed
+    elif section == "hosts":
+        data["hosts"] = parsed
+    elif section == "dns":
+        if parsed:
+            for key in _DNS_KEYS:
+                if key in parsed:
+                    data[key] = parsed[key]
+        else:
+            for key in _DNS_KEYS:
+                data.pop(key, None)
+    else:
+        raise ValueError(f"Unknown YAML section: {section!r}")

hbd/server/dns.py

@@ -4,6 +4,9 @@ from __future__ import annotations
 from subprocess import Popen, PIPE, STDOUT
 from typing import Optional
 import asyncio
+import logging
+
+logger = logging.getLogger(__name__)
 
 
 def create_nsupdate_payload(
@@ -123,7 +126,6 @@ async def dns_update_worker(
                 pass
             continue
 
-        m = f"changed address to {addr}"
         for dyndomain in cfg.get("dyndomains", []):
             err = await loop.run_in_executor(
                 None,
@@ -135,28 +137,29 @@ async def dns_update_worker(
                 cfg.get("rndc_key", "/etc/dhcpc/rndc-key"),
             )
             if err:
-                m += f", DNS update failed: {err}"
+                m = f"DNS update failed for {addr} ({dyndomain}): {err}"
                 logger.error("DNS update failed for %s: %s", name, err)
+                if log:
+                    try:
+                        await loop.run_in_executor(None, log, name, "ERROR", m)
+                    except Exception:
+                        pass
             else:
-                m += ", DNS updated."
+                m = f"DNS updated {name}.dy.{dyndomain} → {addr}"
+                if log:
+                    try:
+                        await loop.run_in_executor(None, log, name, "INFO", m)
+                    except Exception:
+                        pass
+
+        if not cfg.get("dyndomains"):
+            logger.warning("DNS update triggered for %s but no dyndomains configured", name)
 
         try:
             dnsq.task_done()
         except Exception:
             pass
 
-        if log:
-            try:
-                await loop.run_in_executor(None, log, name, m)
-            except Exception:
-                pass
-
-    if log:
-        try:
-            await loop.run_in_executor(None, log, None, "dns_update_worker exiting")
-        except Exception:
-            pass
-
 
 def start_dns_worker(
     hbdclass,

hbd/server/flap.py

@@ -0,0 +1,119 @@
+"""Flap detection: silence checks that toggle faster than they are useful.
+
+A ``(host, service)`` pair is *flapping* once it exceeds ``flap_count``
+WARNING/CRITICAL notifications within ``flap_interval`` minutes.  The
+notification that trips the threshold is delivered with ``FLAP_MARKER``
+appended; every notification for that pair afterwards is dropped.  The state
+clears silently ``flap_interval`` minutes after a RECOVER, provided no further
+WARNING/CRITICAL arrived in the meantime.
+
+Only outbound notifications are affected — ``notify.eventlog`` keeps recording
+every event, so the journal and ``/log`` retain the full history of the flap.
+
+State lives here at module level rather than on ``Host`` so it is never
+pickled: a restart starts every check with a clean slate.
+"""
+
+import logging
+import time
+
+logger = logging.getLogger(__name__)
+
+FLAP_MARKER = "Now flapping!! No more messages!"
+
+# Actions returned by observe()
+PASS = "pass"           # deliver unchanged
+TRIP = "trip"           # deliver with FLAP_MARKER appended
+SUPPRESS = "suppress"   # drop
+
+_count = 0      # flap_count: alerts within the window needed to trip
+_window = 0.0   # flap_interval in seconds
+
+# {(host, service): {"events": [ts, ...], "flapping": bool, "ok_since": float|None}}
+_state: dict = {}
+
+
+def setup(cfg) -> None:
+    """Read flap_count / flap_interval from *cfg* (also called on reload)."""
+    global _count, _window
+    _count = int(cfg.get("flap_count", 0) or 0)
+    _window = float(cfg.get("flap_interval", 0) or 0) * 60.0
+
+
+def _enabled() -> bool:
+    return _count > 0 and _window > 0
+
+
+def _label(key) -> str:
+    host, service = key
+    return f"{host}/{service}" if service else host
+
+
+def _sweep(st: dict, now: float) -> bool:
+    """Clear the flapping flag once the post-RECOVER quiet window has elapsed."""
+    if st["flapping"] and st["ok_since"] is not None and now - st["ok_since"] >= _window:
+        st["flapping"] = False
+        st["ok_since"] = None
+        st["events"].clear()
+        return True
+    return False
+
+
+def observe(host: str, service: str, level: str) -> str:
+    """Record a notification for *host*/*service* and return what to do with it.
+
+    Returns PASS, TRIP or SUPPRESS.
+    """
+    if not _enabled():
+        return PASS
+
+    now = time.time()
+    key = (host or "", service or "")
+    level = (level or "").upper()
+    st = _state.get(key)
+    if st is not None and _sweep(st, now):
+        logger.info("flapping cleared for %s", _label(key))
+
+    if level in ("WARNING", "CRITICAL"):
+        if st is None:
+            st = _state[key] = {"events": [], "flapping": False, "ok_since": None}
+        # An alert inside the quiet window means it never settled.
+        st["ok_since"] = None
+        st["events"] = [t for t in st["events"] if now - t < _window]
+        st["events"].append(now)
+        if st["flapping"]:
+            return SUPPRESS
+        if len(st["events"]) > _count:
+            st["flapping"] = True
+            logger.info(
+                "flapping detected for %s (%d alerts in %.0f min)",
+                _label(key), len(st["events"]), _window / 60,
+            )
+            return TRIP
+        return PASS
+
+    if st is None or not st["flapping"]:
+        return PASS
+    if level == "RECOVER":
+        # Starts the quiet window; the state ends silently when it elapses.
+        st["ok_since"] = now
+    return SUPPRESS
+
+
+def flapping_services(host: str) -> list:
+    """Return the services of *host* that are currently flapping.
+
+    An empty string in the result means the host itself (connectivity, boot,
+    shutdown) rather than a named service.
+    """
+    if not _enabled():
+        return []
+    now = time.time()
+    flapping = []
+    for (h, service), st in _state.items():
+        if h != host:
+            continue
+        _sweep(st, now)
+        if st["flapping"]:
+            flapping.append(service)
+    return sorted(flapping)

hbd/server/hbdclass.py

@@ -95,7 +95,7 @@ class Connection:
         if not Null:
             d["addr"] = self.addr
             if self.rtts[-1]:
-                d["rtt"] = "%0.1f" % self.rtts[-1]
+                d["rtt"] = "%d" % round(self.rtts[-1])
             elif self.state == Connection.UNKNOWN:
                 d["rtt"] = ""
             else:
@@ -297,13 +297,24 @@ class Host:
         self.plugin_retention = 100  # Keep last N samples per plugin
         # Alert state tracking: {metric_path: AlertState}
         self.alert_states = {}
+        # Stale-data timers: {plugin_name: asyncio.TimerHandle}
+        self.plugin_timers = {}
         # User access control
         self.owner: str | None = None       # username of owner
         self.managers: list = []            # usernames with manager role
         self.monitors: list = []            # usernames with monitor role
 
+    def __getstate__(self):
+        """Prepare Host for pickling by excluding non-serializable timer objects."""
+        state = self.__dict__.copy()
+        # asyncio TimerHandles (and their lambda callbacks) can't be pickled.
+        # They're recreated when the next PLG arrives after unpickling.
+        state['plugin_timers'] = {}
+        return state
+
     def statedict(self):
         d = {}
+        d["raw_name"] = self.name
         d["name"] = self.name
         if self.dyn:
             d["name"] += "*"
@@ -364,7 +375,7 @@ class Host:
     def stateinfo(self):
         ddict = {}
         for d in self.__dict__:
-            if d in ["alert_states", "plugin_data"]:
+            if d in ["alert_states", "plugin_data", "plugin_timers"]:
                 continue
             if d == "connections":
                 cl = []
@@ -417,6 +428,10 @@ class Host:
         ddict["alert_critical_unacked"] = critical_unacked
         ddict["alert_critical_acked"] = critical_acked
 
+        # Flap detection state (module-level in flap.py, never pickled)
+        from . import flap
+        ddict["flapping"] = flap.flapping_services(self.name)
+
         # User access
         ddict["owner"] = getattr(self, "owner", None)
         ddict["managers"] = list(getattr(self, "managers", []))
@@ -482,6 +497,8 @@ class Host:
             self.managers = []
         if not hasattr(self, "monitors"):
             self.monitors = []
+        if not hasattr(self, "plugin_timers"):
+            self.plugin_timers = {}
 
         pass
 
@@ -541,6 +558,34 @@ class Host:
         """
         return self.plugin_data
 
+    def reset_plugin_timer(self, plugin_name, timeout_seconds, callback):
+        """Reset the stale-data timer for a plugin.
+
+        If no new PLG data arrives within timeout_seconds, callback(host, plugin_name)
+        is called so the caller can clear history and alerts.
+        """
+        import asyncio
+        existing = self.plugin_timers.get(plugin_name)
+        if existing and not existing.cancelled():
+            existing.cancel()
+
+        async def _fire():
+            await callback(self, plugin_name)
+
+        try:
+            loop = asyncio.get_event_loop()
+            self.plugin_timers[plugin_name] = loop.call_later(
+                timeout_seconds, lambda: asyncio.create_task(_fire())
+            )
+        except RuntimeError:
+            pass
+
+    def cancel_plugin_timer(self, plugin_name):
+        """Cancel the stale timer for a plugin, if any."""
+        handle = self.plugin_timers.pop(plugin_name, None)
+        if handle and not handle.cancelled():
+            handle.cancel()
+
     # ------------------------------------------------------------------
     # User-role helpers
     # ------------------------------------------------------------------

hbd/server/journal.py

@@ -11,7 +11,7 @@ import os
 import asyncio
 from datetime import datetime
 from pathlib import Path
-from typing import Dict, Any, Optional
+from typing import Dict, Any, Optional, Callable, Iterable, List, Tuple, Union
 
 logger = logging.getLogger(__name__)
 
@@ -153,6 +153,31 @@ class MessageJournal:
             except Exception as e:
                 logger.error(f"Error writing to journal: {e}")
 
+    async def log_event(self, event: Dict[str, Any]):
+        """Write a caller-provided dict verbatim as one JSONL line (with rotation)."""
+        if not self.enabled or not self._initialized:
+            return
+        async with self._lock:
+            try:
+                line = json.dumps(event, separators=(',', ':')) + '\n'
+                nbytes = len(line.encode('utf-8'))
+                if self._current_size + nbytes > self.max_size:
+                    await self._rotate()
+                if self._file_handle:
+                    self._file_handle.write(line)
+                    self._file_handle.flush()
+                    self._current_size += nbytes
+            except Exception as e:
+                logger.error(f"Error writing event to journal: {e}")
+
+    async def backfill(self, events: Iterable[Dict[str, Any]]):
+        """One-time seed: write *events* only if the journal file is currently empty."""
+        if not self.enabled or not self._initialized or self._current_size > 0:
+            return
+        for ev in events:
+            if isinstance(ev, dict):
+                await self.log_event(ev)
+
     async def _rotate(self):
         """
         Rotate the journal file.
@@ -309,6 +334,88 @@ class MessageJournal:
         }
 
 
+def _iter_journal_events(journal_dir: Union[str, Path], journal_file: str) -> Iterable[Dict[str, Any]]:
+    """Yield event dicts from the journal files, newest event first.
+
+    Reads the current file, then rotated backups newest-to-oldest (backup
+    names embed rotation timestamps, so filename sort is chronological).
+    """
+    dirp = Path(journal_dir)
+    files = [dirp / journal_file]
+    files.extend(sorted(dirp.glob(journal_file + '.*'), reverse=True))
+    for f in files:
+        if not f.is_file():
+            continue
+        try:
+            lines = f.read_text(encoding='utf-8', errors='replace').splitlines()
+        except OSError as e:
+            logger.warning(f"Cannot read journal file {f}: {e}")
+            continue
+        for line in reversed(lines):
+            try:
+                ev = json.loads(line)
+            except ValueError:
+                continue
+            if isinstance(ev, dict):
+                yield ev
+
+
+def filter_events(
+    events: Iterable[Dict[str, Any]],
+    limit: int = 100,
+    before: Optional[float] = None,
+    host: Optional[str] = None,
+    level: Optional[str] = None,
+    q: Optional[str] = None,
+    predicate: Optional[Callable[[Dict[str, Any]], bool]] = None,
+) -> Tuple[List[Dict[str, Any]], bool]:
+    """Filter an iterable of event dicts already ordered newest-first.
+
+    Returns (events, more): up to *limit* matching events, and whether at
+    least one further matching event exists beyond the limit.
+    """
+    host_l = host.lower() if host else None
+    level_l = level.lower() if level else None
+    q_l = q.lower() if q else None
+    out: List[Dict[str, Any]] = []
+    for ev in events:
+        if not isinstance(ev, dict):
+            continue
+        ts = ev.get('ts')
+        if before is not None and (not isinstance(ts, (int, float)) or ts >= before):
+            continue
+        if host_l and host_l not in str(ev.get('host') or '').lower():
+            continue
+        if level_l and str(ev.get('level') or '').lower() != level_l:
+            continue
+        if q_l and q_l not in str(ev.get('message') or '').lower():
+            continue
+        if predicate is not None and not predicate(ev):
+            continue
+        if len(out) >= limit:
+            return out, True
+        out.append(ev)
+    return out, False
+
+
+def read_events(
+    journal_dir: Union[str, Path],
+    journal_file: str = 'events.journal',
+    *,
+    limit: int = 100,
+    before: Optional[float] = None,
+    host: Optional[str] = None,
+    level: Optional[str] = None,
+    q: Optional[str] = None,
+    predicate: Optional[Callable[[Dict[str, Any]], bool]] = None,
+) -> Tuple[List[Dict[str, Any]], bool]:
+    """Read filtered events newest-first from the events journal files."""
+    return filter_events(
+        _iter_journal_events(journal_dir, journal_file),
+        limit=limit, before=before, host=host, level=level, q=q, predicate=predicate,
+    )
+
+
 # Global journal instance
 _journal_instance: Optional[MessageJournal] = None
 
@@ -340,3 +447,28 @@ async def log_message(msg: Dict[str, Any], addr: tuple, timestamp: Optional[floa
     """
     journal = get_journal()
     await journal.log_message(msg, addr, timestamp)
+
+
+# Global events journal instance (human-readable event log, written by notify.eventlog)
+_events_journal_instance: Optional[MessageJournal] = None
+
+
+def get_events_journal(config: Optional[Dict[str, Any]] = None) -> MessageJournal:
+    """Get or create the global events journal instance.
+
+    Uses the events_journal_* config keys; shares journal_dir and
+    journal_enabled with the raw-datagram journal.
+    """
+    global _events_journal_instance
+    if _events_journal_instance is None:
+        cfg = config or {}
+        _events_journal_instance = MessageJournal(
+            {
+                'journal_dir': cfg.get('journal_dir', '/var/log/heartbeat'),
+                'journal_file': cfg.get('events_journal_file', 'events.journal'),
+                'journal_max_size': cfg.get('events_journal_max_size', 10 * 1024 * 1024),
+                'journal_max_backups': cfg.get('events_journal_max_backups', 10),
+                'journal_enabled': cfg.get('journal_enabled', True),
+            }
+        )
+    return _events_journal_instance

hbd/server/main.py

@@ -78,9 +78,7 @@ async def reload_configuration(config_obj, config_path, components):
         True if reload succeeded, False otherwise
     """
     try:
-        logger.info("=" * 60)
         logger.info("Starting configuration reload...")
-        logger.info("=" * 60)
         
         # Reload config file
         new_config = await config_obj.reload(config_path)
@@ -101,9 +99,10 @@ async def reload_configuration(config_obj, config_path, components):
             access = config_mod.get_host_access(new_config, hostname)
             host.apply_access(access["owner"], access["managers"], access["monitors"])
 
-        # Reload threshold checker
+        # Reload threshold checker and prune alerts orphaned by the new config
         if 'threshold_checker' in components:
             components['threshold_checker'].reload(new_config)
+            components['threshold_checker'].purge_stale_alerts(hbdclass)
         
         # Note: Changes to the following require restart:
         # - hb_port, hbd_port, ws_port (already bound)
@@ -114,13 +113,11 @@ async def reload_configuration(config_obj, config_path, components):
         # These are reloadable and effective immediately:
         # - notification_channels
         # - threshold_configs
-        # - hosts (watchhosts, dyndnshosts, notification_channels)
+        # - hosts (watchhosts, dyndns, notification_channels)
         # - grace period (used on next heartbeat)
         # - debug/verbose flags (used on next message)
         
-        logger.info("=" * 60)
         logger.info("Configuration reload completed successfully")
-        logger.info("=" * 60)
         return True
         
     except Exception as e:
@@ -168,6 +165,15 @@ async def _run_async(config, config_path=None):
     msg_journal = journal_mod.get_journal(config)
     await msg_journal.initialize()
 
+    # Initialize events journal (human-readable event log for the /log page)
+    events_journal = journal_mod.get_events_journal(config)
+    await events_journal.initialize()
+    if data.msgs:
+        # One-time seed on upgrade: only writes when the journal file is empty
+        await events_journal.backfill(data.msgs)
+
+    notify_mod.eventlog(None, "INFO", f"hbd version {__version__} starting up")
+
     # Initialize threshold checker
     threshold_checker = threshold_mod.ThresholdChecker(
         config=config,
@@ -241,6 +247,13 @@ async def _run_async(config, config_path=None):
     )
     udp.restore_connection_timers(hbdclass, restore_ctx)
 
+    # Drop alert states that no longer have a matching threshold (stale after
+    # upgrade or config change between runs).
+    threshold_checker.purge_stale_alerts(hbdclass)
+
+    async def _http_reload_callback():
+        await reload_configuration(config, config_path, components)
+
     # HTTP server (asyncio-based via aiohttp)
     try:
         http_task = asyncio.create_task(
@@ -250,9 +263,11 @@ async def _run_async(config, config_path=None):
                 config=config,
                 hbdclass=hbdclass,
                 tcss=None,
+                threshold_checker=threshold_checker,
                 verbose=config.get("verbose", False),
                 get_now=lambda: time.time(),
                 VER="",
+                reload_callback=_http_reload_callback,
             )
         )
         logger.info(
@@ -373,6 +388,26 @@ async def _run_async(config, config_path=None):
         except Exception as e:
             logger.warning("Error closing message journal: %s", e)
 
+        # Journal the shutdown event now; run()'s shutdown eventlog fires after this journal closes
+        try:
+            await events_journal.log_event(
+                {
+                    "ts": time.time(),
+                    "host": None,
+                    "level": "INFO",
+                    "service": None,
+                    "message": f"hbd version {__version__} shutdown",
+                }
+            )
+        except Exception as e:
+            logger.debug("Could not journal shutdown event: %s", e)
+
+        # Close events journal
+        try:
+            await events_journal.close()
+        except Exception as e:
+            logger.warning("Error closing events journal: %s", e)
+
         # Signal DNS worker to exit and await it
         try:
             if "dns_task" in locals() and dns_task:
@@ -416,7 +451,6 @@ def load_pickled_hosts(config, hbdclass):
     pickfile = config.get("pickfile", "hbd.pickle")
     dyndnshosts = config_mod.get_dyndnshosts(config)
     watchhosts = config_mod.get_watchhosts(config)
-    drophosts = config.get("drophosts", [])
     if 1 and os.path.exists(pickfile):
         if config.get("verbose", False):
             logger.info("opening pickls %s", pickfile)
@@ -442,9 +476,6 @@ def load_pickled_hosts(config, hbdclass):
             hbdclass.Host.hosts[h].apply_access(
                 access["owner"], access["managers"], access["monitors"]
             )
-        for h in drophosts:
-            if h in hbdclass.Host.hosts:
-                del hbdclass.Host.hosts[h]
         if config.get("verbose", False):
             logger.info("%s pickled hosts loaded", len(hbdclass.Host.hosts))
     else:
@@ -469,6 +500,8 @@ def run(config, config_path=None):
     if config.get("debug", 0) > 0:
         log_level = logging.DEBUG
     logging.basicConfig(level=log_level)
+    if not config.get("debug", 0):
+        logging.getLogger("aiohttp.access").propagate = False
     load_pickled_hosts(config, hbdclass)
 
     notify_mod.initlog(logfile=config.get("logfile", "messages.log"))
@@ -483,8 +516,6 @@ def run(config, config_path=None):
         except Exception as e:
             logger.warning("Failed to write pidfile %s: %s", pidfile, e)
 
-    eventlog(None, "INFO", f"hbd version {__version__} starting up")
-    
     if config_path:
         logger.info(f"Config file: {config_path} (reload with SIGHUP)")
     else:

hbd/server/notify.py

@@ -25,6 +25,7 @@ from dataclasses import dataclass, field
 from typing import Optional
 
 from . import data
+from . import flap as flap_mod
 from . import ws as ws_mod
 
 logger = logging.getLogger(__name__)
@@ -33,6 +34,7 @@ msg_to_websockets = ws_mod.broadcast
 
 # Module-level state set via setup()
 _config: dict = {}
+_loop: Optional[asyncio.AbstractEventLoop] = None
 
 # Tracks which channels fired a WARNING/CRITICAL per host.
 # {host_name: set of channel_names}  — used to route RECOVER to the same channels.
@@ -62,6 +64,7 @@ class Notification:
     body: str           # detail message
     level: str          # RECOVER | WARNING | CRITICAL | INFO
     url: str = ""       # link to plugin metrics page
+    service: str = ""   # flap-detection key within the host ("" = the host itself)
 
 
 # ---------------------------------------------------------------------------
@@ -70,14 +73,18 @@ class Notification:
 
 def setup(cfg: dict, loop: Optional[asyncio.AbstractEventLoop] = None):
     """Initialize notifier from configuration dict."""
-    global _config
+    global _config, _loop
     _config = dict(cfg)
+    flap_mod.setup(_config)
+    if loop is not None:
+        _loop = loop
 
 
 def reload_config(cfg: dict):
     """Reload notification configuration on SIGHUP."""
     global _config
     _config = dict(cfg)
+    flap_mod.setup(_config)
     logger.info("Notification configuration reloaded")
 
 
@@ -106,11 +113,23 @@ def closelog():
 
 def eventlog(host, lvl, m, service=None):
     ts = time.time()
+    msg = {
+        "ts": ts,
+        "host": host or None,
+        "level": lvl,
+        "service": service,
+        "message": m,
+    }
+    data.msgs.append(msg)
+    # Cap the in-memory buffer so it doesn't grow without bound; this list is
+    # replayed to every dashboard client on connect and persisted in the pickle.
+    cap = _config.get("msg_buffer_size", 500)
+    if cap and len(data.msgs) > cap:
+        del data.msgs[:-cap]
     s = f"{time.strftime('%Y-%m-%d %H:%M:%S', time.localtime(ts))} {lvl} "
     if host:
         s += f"{host} "
     s += m
-    data.msgs.append(s)
     logger.info(s)
     if logf:
         try:
@@ -118,7 +137,22 @@ def eventlog(host, lvl, m, service=None):
             logf.flush()
         except Exception as e:
             logger.warning("failed to write to logfile: %s", e)
-    msg_to_websockets("message", s)
+    msg_to_websockets("message", msg)
+    _journal_event(msg)
+
+
+def _journal_event(msg: dict):
+    """Schedule an async write of *msg* to the events journal (no-op without a loop)."""
+    if _loop is None or not _loop.is_running():
+        return
+    from . import journal as journal_mod
+    ej = journal_mod._events_journal_instance
+    if ej is None or not ej.enabled:
+        return
+    try:
+        asyncio.run_coroutine_threadsafe(ej.log_event(msg), _loop)
+    except Exception as e:
+        logger.warning("failed to schedule events journal write: %s", e)
 
 
 # ---------------------------------------------------------------------------
@@ -133,10 +167,14 @@ def _send_pushover(channel_cfg: dict, notif: Notification) -> bool:
     if not token or not user:
         logger.warning("pushover: missing token or user")
         return False
-    params: dict = {"token": token, "user": user, "title": notif.title, "message": notif.body}
+    body = "%s: %s" % (notif.title, notif.body)
+    title = ""
+    params: dict = {"token": token, "user": user, "title": title, "message": body}
+    if channel_cfg.get("sound"):
+        params["sound"] = channel_cfg["sound"]
     if notif.url:
         params["url"] = notif.url
-        params["url_title"] = "Plugin metrics"
+        params["url_title"] = "Heartbeat"
     conn = http.client.HTTPSConnection("api.pushover.net:443")
     try:
         conn.request(
@@ -209,7 +247,7 @@ def _send_mattermost(channel_cfg: dict, notif: Notification) -> bool:
         return False
     text = f"**{notif.title}**\n{notif.body}"
     if notif.url:
-        text += f"\n[Plugin metrics]({notif.url})"
+        text += f"\n[Plugin metrics] {notif.url}"
     ses = {"url": host, "scheme": "http", "basepath": "/api/v4", "port": 8065}
     mm = Driver(ses)
     payload: dict = {"text": text, "channel": channel, "username": channel_cfg.get("username", "hbd")}
@@ -357,6 +395,9 @@ _TIMEOUT = 15  # seconds per channel send
 
 async def _dispatch_to_channel(channel_name: str, channel_cfg: dict, notif: Notification) -> bool:
     """Send *notif* to a single named channel, honouring min_level."""
+    # Strip ownership metadata — notifier drivers only need delivery credentials.
+    channel_cfg = {k: v for k, v in channel_cfg.items() if k not in ("owner", "private")}
+
     level = notif.level.upper()
     if level != "RECOVER":
         min_level = channel_cfg.get("min_level", "WARNING").upper()
@@ -392,7 +433,7 @@ def _build_url(host_name: str) -> str:
     base_url = _config.get("base_url", "").rstrip("/")
     if not base_url:
         return ""
-    return f"{base_url}/plugins#{host_name}"
+    return f"{base_url}/alerts?filter={host_name}"
 
 
 async def send_notification(host_name: str, notif: Notification) -> dict:
@@ -402,11 +443,22 @@ async def send_notification(host_name: str, notif: Notification) -> dict:
     notification_channels, and dispatches.  Silently does nothing if
     no users are configured.
 
+    Flap detection runs first: once *host_name*/*notif.service* is flapping the
+    notification is dropped, and the one that trips the state carries
+    ``flap.FLAP_MARKER``.
+
     Returns a dict of {channel_name: bool} results.
     """
     from . import users as users_mod
     from . import hbdclass
 
+    action = flap_mod.observe(host_name, notif.service, notif.level)
+    if action == flap_mod.SUPPRESS:
+        logger.debug("flapping: suppressed %s notification for %s", notif.level, host_name)
+        return {}
+    if action == flap_mod.TRIP:
+        notif.body = f"{notif.body} {flap_mod.FLAP_MARKER}"
+
     if not users_mod.users_enabled():
         return {}
 

hbd/server/oauth.py

@@ -0,0 +1,254 @@
+"""OAuth2 provider support.
+
+Config shape (in ~/.hb.yaml):
+
+    oauth:
+      my-gitea:                          # route slug → /login/oauth/my-gitea
+        type: gitea                      # "gitea" | "github" | "nextcloud"
+                                         # omit type to default to "gitea"
+        url: https://git.example.com     # required for gitea and nextcloud
+        client_id: <client-id>
+        client_secret: <client-secret>
+        label: "Work Gitea"              # optional display name on login button
+        logo: https://example.com/logo.png  # optional logo URL
+
+      github:
+        type: github
+        client_id: <client-id>
+        client_secret: <client-secret>
+
+      nextcloud:
+        type: nextcloud
+        url: https://cloud.example.com
+        client_id: <client-id>
+        client_secret: <client-secret>
+
+Register the OAuth app with each provider and set the redirect URI to:
+  https://<hbd-host>/login/oauth/<name>/callback
+"""
+
+import logging
+import secrets
+import time
+import urllib.parse
+from dataclasses import dataclass
+
+import aiohttp
+
+logger = logging.getLogger(__name__)
+
+STATE_TTL = 600  # 10 minutes
+
+# state_token -> expiry timestamp
+_states: dict[str, float] = {}
+
+
+def make_state() -> str:
+    """Generate a CSRF state token, store it with TTL, and return it."""
+    _purge_states()
+    token = secrets.token_hex(32)
+    _states[token] = time.time() + STATE_TTL
+    return token
+
+
+def validate_state(state: str) -> bool:
+    """Return True if *state* is known and unexpired; always removes it."""
+    expiry = _states.pop(state, None)
+    if expiry is None:
+        return False
+    return time.time() < expiry
+
+
+def _purge_states() -> None:
+    """Remove all expired CSRF state tokens from the in-memory store."""
+    now = time.time()
+    expired = [k for k, exp in list(_states.items()) if exp < now]
+    for k in expired:
+        del _states[k]
+
+
+class OAuthError(Exception):
+    """Raised when the OAuth2 flow fails for any reason."""
+
+
+PROVIDER_DEFS: dict = {
+    "gitea": {
+        "authorize_url_tmpl": "{url}/login/oauth/authorize",
+        "token_url_tmpl":     "{url}/login/oauth/access_token",
+        "profile_url_tmpl":   "{url}/api/v1/user",
+        "scope":              "user:email",
+        "field_map":          {"username": "login", "full_name": "full_name", "avatar": "avatar_url"},
+        "profile_data_path":  [],
+        "requires_url":       True,
+        "default_label":      "Gitea",
+    },
+    "github": {
+        "authorize_url_tmpl": "https://github.com/login/oauth/authorize",
+        "token_url_tmpl":     "https://github.com/login/oauth/access_token",
+        "profile_url_tmpl":   "https://api.github.com/user",
+        "scope":              "read:user",
+        "field_map":          {"username": "login", "full_name": "name", "avatar": "avatar_url"},
+        "profile_data_path":  [],
+        "requires_url":       False,
+        "default_label":      "GitHub",
+    },
+    "nextcloud": {
+        "authorize_url_tmpl": "{url}/apps/oauth2/authorize",
+        "token_url_tmpl":     "{url}/apps/oauth2/api/v1/token",
+        "profile_url_tmpl":   "{url}/ocs/v2.php/cloud/user?format=json",
+        "scope":              "",
+        "field_map":          {"username": "id", "full_name": "display-name", "avatar": None},
+        "profile_data_path":  ["ocs", "data"],
+        "requires_url":       True,
+        "default_label":      "Nextcloud",
+    },
+}
+
+
+@dataclass
+class ResolvedProvider:
+    """A fully resolved OAuth2 provider instance, ready to use."""
+    name: str
+    type: str
+    label: str
+    logo: str
+    authorize_url: str
+    token_url: str
+    profile_url: str
+    scope: str
+    client_id: str
+    client_secret: str
+    field_map: dict
+    profile_data_path: list
+
+
+def get_providers(config: dict) -> list[ResolvedProvider]:
+    """Return a ResolvedProvider for every valid entry in config['oauth'].
+
+    Entries with missing required fields or unknown types are skipped with
+    a warning log.  Order follows config declaration order.
+    """
+    result = []
+    oauth_cfg = config.get("oauth", {})
+    if not isinstance(oauth_cfg, dict):
+        return result
+    for name, entry in oauth_cfg.items():
+        if not isinstance(entry, dict):
+            continue
+        provider_type = entry.get("type", "gitea")
+        defn = PROVIDER_DEFS.get(provider_type)
+        if defn is None:
+            logger.warning("OAuth: unknown provider type %r for %r, skipping", provider_type, name)
+            continue
+        client_id = entry.get("client_id", "")
+        client_secret = entry.get("client_secret", "")
+        if not client_id or not client_secret:
+            logger.warning("OAuth: %r missing client_id or client_secret, skipping", name)
+            continue
+        url = entry.get("url", "").rstrip("/")
+        if defn["requires_url"] and not url:
+            logger.warning("OAuth: %r requires url but none configured, skipping", name)
+            continue
+        label = entry.get("label") or defn["default_label"]
+        logo = entry.get("logo", "")
+        result.append(ResolvedProvider(
+            name=name,
+            type=provider_type,
+            label=label,
+            logo=logo,
+            authorize_url=defn["authorize_url_tmpl"].format(url=url),
+            token_url=defn["token_url_tmpl"].format(url=url),
+            profile_url=defn["profile_url_tmpl"].format(url=url),
+            scope=defn["scope"],
+            client_id=client_id,
+            client_secret=client_secret,
+            field_map=dict(defn["field_map"]),
+            profile_data_path=list(defn["profile_data_path"]),
+        ))
+    return result
+
+
+def is_enabled(config: dict) -> bool:
+    """Return True when at least one OAuth provider is fully configured."""
+    return bool(get_providers(config))
+
+
+def build_auth_url(provider: ResolvedProvider, state: str, redirect_uri: str) -> str:
+    """Return the provider's OAuth2 authorization URL to redirect the browser to."""
+    params: dict = {
+        "client_id": provider.client_id,
+        "redirect_uri": redirect_uri,
+        "response_type": "code",
+        "state": state,
+    }
+    if provider.scope:
+        params["scope"] = provider.scope
+    return f"{provider.authorize_url}?{urllib.parse.urlencode(params)}"
+
+
+async def exchange_code(provider: ResolvedProvider, code: str, redirect_uri: str) -> str:
+    """Exchange an authorization *code* for an access token.
+
+    Returns the access token string.  Raises OAuthError on any failure.
+    """
+    payload = {
+        "client_id": provider.client_id,
+        "client_secret": provider.client_secret,
+        "code": code,
+        "grant_type": "authorization_code",
+        "redirect_uri": redirect_uri,
+    }
+    timeout = aiohttp.ClientTimeout(total=10)
+    try:
+        async with aiohttp.ClientSession(timeout=timeout) as session:
+            async with session.post(
+                provider.token_url,
+                json=payload,
+                headers={"Accept": "application/json"},
+            ) as resp:
+                if resp.status != 200:
+                    text = await resp.text()
+                    raise OAuthError(f"Token exchange failed ({resp.status}): {text}")
+                data = await resp.json()
+                token = data.get("access_token")
+                if not token:
+                    raise OAuthError(f"No access_token in response: {data}")
+    except aiohttp.ClientError as exc:
+        raise OAuthError(f"Token exchange network error: {exc}") from exc
+    return token
+
+
+async def fetch_user(provider: ResolvedProvider, token: str) -> dict:
+    """Fetch the authenticated user's profile from the provider.
+
+    Returns a dict with keys: login, full_name, avatar_url.
+    Raises OAuthError on any failure.
+    """
+    timeout = aiohttp.ClientTimeout(total=10)
+    try:
+        async with aiohttp.ClientSession(timeout=timeout) as session:
+            async with session.get(
+                provider.profile_url,
+                headers={
+                    "Authorization": f"Bearer {token}",
+                    "Accept": "application/json",
+                },
+            ) as resp:
+                if resp.status != 200:
+                    text = await resp.text()
+                    raise OAuthError(f"User fetch failed ({resp.status}): {text}")
+                data = await resp.json()
+    except aiohttp.ClientError as exc:
+        raise OAuthError(f"User fetch network error: {exc}") from exc
+
+    try:
+        for key in provider.profile_data_path:
+            data = data.get(key, {})
+        avatar_field = provider.field_map.get("avatar")
+        return {
+            "login":      data.get(provider.field_map["username"], ""),
+            "full_name":  data.get(provider.field_map["full_name"], ""),
+            "avatar_url": data.get(avatar_field, "") if avatar_field else "",
+        }
+    except AttributeError:
+        raise OAuthError(f"Unexpected profile response structure from {provider.type}")

hbd/server/settings.py

@@ -21,19 +21,73 @@ editable    bool  Reserved for future use — currently always False
 sensitive   bool  True when the raw value must never be shown
 """
 
+from . import config_access
+
 # Credential field names that should always be masked.
 _SECRET_KEYS = frozenset({
     "password", "token", "user_key", "api_key", "secret",
-    "smtp_password", "smtp_user",
+    "smtp_password", "smtp_user", "api_password", "access_token",
 })
 
-_CHANNEL_TYPE_LABELS = {
+CHANNEL_TYPE_SCHEMAS = {
-    "pushover":   "Pushover",
+    "pushover": {
-    "email":      "E-mail",
+        "label": "Pushover",
-    "signal":     "Signal",
+        "fields": [
-    "mattermost": "Mattermost",
+            {"key": "token",  "label": "App token",  "type": "secret", "required": True},
+            {"key": "user",   "label": "User key",   "type": "secret", "required": True},
+            {"key": "sound",  "label": "Sound",      "type": "text",   "required": False},
+        ],
+    },
+    "email": {
+        "label": "E-mail",
+        "fields": [
+            {"key": "recipients",    "label": "Recipients (comma-separated)", "type": "list",   "required": True},
+            {"key": "sender",        "label": "From address",                 "type": "text",   "required": True},
+            {"key": "smtp_server",   "label": "SMTP server",                  "type": "text",   "required": True},
+            {"key": "smtp_port",     "label": "SMTP port",                    "type": "port",   "required": False},
+            {"key": "smtp_user",     "label": "SMTP username",                "type": "text",   "required": False},
+            {"key": "smtp_password", "label": "SMTP password",                "type": "secret", "required": False},
+        ],
+    },
+    "signal": {
+        "label": "Signal",
+        "fields": [
+            {"key": "user",      "label": "Sender number",    "type": "text", "required": True},
+            {"key": "recipient", "label": "Recipient number", "type": "text", "required": True},
+            {"key": "cli_path",  "label": "signal-cli path",  "type": "text", "required": False},
+        ],
+    },
+    "matrix": {
+        "label": "Matrix",
+        "fields": [
+            {"key": "homeserver",   "label": "Homeserver URL", "type": "text",   "required": True},
+            {"key": "access_token", "label": "Access token",   "type": "secret", "required": True},
+            {"key": "room_id",      "label": "Room ID",        "type": "text",   "required": True},
+        ],
+    },
+    "sms_voipms": {
+        "label": "SMS (voip.ms)",
+        "fields": [
+            {"key": "api_user",     "label": "API username", "type": "text",   "required": True},
+            {"key": "api_password", "label": "API password", "type": "secret", "required": True},
+            {"key": "did",          "label": "DID (from)",   "type": "text",   "required": True},
+            {"key": "dst",          "label": "Destination",  "type": "text",   "required": True},
+        ],
+    },
+    "mattermost": {
+        "label": "Mattermost",
+        "fields": [
+            {"key": "host",     "label": "Host",          "type": "text",   "required": True},
+            {"key": "token",    "label": "Webhook token", "type": "secret", "required": True},
+            {"key": "channel",  "label": "Channel",       "type": "text",   "required": True},
+            {"key": "username", "label": "Bot username",  "type": "text",   "required": False},
+            {"key": "icon",     "label": "Icon URL",      "type": "text",   "required": False},
+        ],
+    },
 }
 
+_CHANNEL_TYPE_LABELS = {k: v["label"] for k, v in CHANNEL_TYPE_SCHEMAS.items()}
+
 
 def _mask(value):
     """Return a masked placeholder for sensitive values."""
@@ -88,9 +142,14 @@ def _sanitize_channel(name, cfg):
 # Public API
 # ---------------------------------------------------------------------------
 
-def get_settings_sections(config: dict) -> list:
+def get_settings_sections(config: dict, threshold_checker=None, user=None) -> list:
     """Return ordered list of setting sections for the settings page.
 
+    *user* is an object with ``username``/``admin`` attributes, or None for
+    the unauthenticated (admin-equivalent) view.  Non-admins get only the
+    channels/hosts/thresholds sections, filtered to global items plus what
+    they own or manage.
+
     Each section:
         {
             "title": str,
@@ -110,6 +169,9 @@ def get_settings_sections(config: dict) -> list:
             "sensitive":   bool,
         }
     """
+    is_admin = user is None or getattr(user, "admin", False)
+    username: str = getattr(user, "username", "") or ""
+
     def field(key, label, ftype, description="", editable=False, sensitive=False):
         raw = config.get(key)
         if sensitive:
@@ -143,14 +205,17 @@ def get_settings_sections(config: dict) -> list:
         }
 
     # ---- Notification channels (complex, built separately) ----------------
+    _METADATA_KEYS = {"type", "owner", "private", "min_level"}
     notif_channels = []
-    for ch_name, ch_cfg in (config.get("notification_channels") or {}).items():
+    for ch_name, ch_cfg in sorted((config.get("notification_channels") or {}).items()):
         if not isinstance(ch_cfg, dict):
             continue
+        if not is_admin and not config_access.user_can_use(ch_cfg, username):
+            continue
         ch_type = ch_cfg.get("type", "")
         fields = []
         for k, v in ch_cfg.items():
-            if k == "type":
+            if k in _METADATA_KEYS:
                 continue
             sensitive = k in _SECRET_KEYS
             fields.append({
@@ -165,60 +230,140 @@ def get_settings_sections(config: dict) -> list:
             "name": ch_name,
             "type": ch_type,
             "type_label": _CHANNEL_TYPE_LABELS.get(ch_type, ch_type.title()),
+            "owner": ch_cfg.get("owner"),
+            "editable": is_admin or ch_cfg.get("owner") == username,
+            "min_level": ch_cfg.get("min_level", "WARNING"),
             "fields": fields,
         })
 
     # ---- Users (show metadata only, never password hashes) ----------------
     users_list = []
-    for username, attrs in (config.get("users") or {}).items():
+    for uname, attrs in (config.get("users") or {}).items():
         if not isinstance(attrs, dict):
             continue
         users_list.append({
-            "username": username,
+            "username": uname,
             "full_name": attrs.get("full_name", ""),
             "admin": bool(attrs.get("admin", False)),
             "avatar": attrs.get("avatar", ""),
             "notification_channels": attrs.get("notification_channels", []),
         })
 
+    # ---- Threshold configurations -----------------------------------------
+    def _tc_to_row(tc):
+        return {
+            "metric": tc.metric_path,
+            "operator": tc.operator.value,
+            "warning": tc.warning,
+            "critical": tc.critical,
+            "hysteresis": tc.hysteresis,
+            "count": tc.count,
+            "enabled": tc.enabled,
+            "display": tc.display or "",
+            "grace": tc.grace,
+        }
+
+    threshold_config_list = []
+    raw_threshold_cfgs = config.get("threshold_configs") or {}
+    if threshold_checker is not None:
+        if threshold_checker.threshold_configs:
+            for cfg_name, cfg_metrics in sorted(threshold_checker.threshold_configs.items()):
+                raw_cfg = raw_threshold_cfgs.get(cfg_name)
+                tc_owner = raw_cfg.get("owner") if isinstance(raw_cfg, dict) else None
+                if not is_admin and tc_owner and tc_owner != username:
+                    continue
+                # For the default config use the merged effective set;
+                # for named overrides use only the explicitly defined metrics
+                # (threshold_raw_configs) so inherited defaults are not repeated.
+                if cfg_name == "default":
+                    display_metrics = cfg_metrics
+                else:
+                    display_metrics = threshold_checker.threshold_raw_configs.get(cfg_name, cfg_metrics)
+                metrics = sorted(
+                    [_tc_to_row(tc) for tc in display_metrics.values()],
+                    key=lambda m: m["metric"],
+                )
+                threshold_config_list.append({
+                    "name": cfg_name,
+                    "metrics": metrics,
+                    "owner": tc_owner,
+                    "editable": is_admin or (tc_owner is not None and tc_owner == username),
+                })
+        elif threshold_checker.thresholds:
+            metrics = sorted(
+                [_tc_to_row(tc) for tc in threshold_checker.thresholds.values()],
+                key=lambda m: m["metric"],
+            )
+            threshold_config_list.append({"name": "default", "metrics": metrics,
+                                          "owner": None, "editable": is_admin})
+
     # ---- Hosts summary ----------------------------------------------------
     hosts_list = []
-    for hname, hcfg in (config.get("hosts") or {}).items():
+    for hname, hcfg in sorted((config.get("hosts") or {}).items()):
         if not isinstance(hcfg, dict):
             continue
+        managers = hcfg.get("managers", [])
+        if isinstance(managers, str):
+            managers = [managers]
+        if not is_admin and hcfg.get("owner") != username and username not in managers:
+            continue
         hosts_list.append({
             "name": hname,
-            "watch": bool(hcfg.get("watch", False)),
+            "watch": bool(hcfg.get("watch", True)),
             "dyndns": bool(hcfg.get("dyndns", False)),
             "owner": hcfg.get("owner", ""),
-            "managers": hcfg.get("managers", []),
+            "is_owner": is_admin or hcfg.get("owner") == username,
+            "managers": managers,
             "monitors": hcfg.get("monitors", []),
-            "threshold_config": hcfg.get("threshold_config", ""),
+            "threshold_configs": (
+                list(v) if isinstance(v := hcfg.get("threshold_config"), list)
+                else ([v] if v else [])
+            ),
             "notification_channels": hcfg.get("notification_channels", []),
         })
 
-    return [
+    # ---- OAuth providers -------------------------------------------------------
+    oauth_providers = []
+    for pname, pattrs in (config.get("oauth") or {}).items():
+        if not isinstance(pattrs, dict):
+            continue
+        cs = pattrs.get("client_secret", "")
+        oauth_providers.append({
+            "name": pname,
+            "type": pattrs.get("type", "gitea"),
+            "url": pattrs.get("url", ""),
+            "client_id": pattrs.get("client_id", ""),
+            "client_secret": "•••" if cs else "",
+            "label": pattrs.get("label", ""),
+            "logo": pattrs.get("logo", ""),
+        })
+
+    sections: list = [
         {
             "id": "network",
             "title": "Network",
             "description": "Ports and bind addresses for all server sockets.",
+            "section_mode": "form",
+            "api_section": "server",
             "fields": [
                 field("hb_port",  "Heartbeat UDP port",  "port",
-                      "UDP port the server listens on for heartbeat datagrams."),
+                      "UDP port the server listens on for heartbeat datagrams.", editable=True),
                 field("hbd_host", "HTTP bind address",   "text",
-                      "Interface to bind the HTTP server to. Empty = all interfaces."),
+                      "Interface to bind the HTTP server to. Empty = all interfaces.", editable=True),
                 field("hbd_port", "HTTP API port",       "port",
-                      "TCP port for the HTTP API and web UI."),
+                      "TCP port for the HTTP API and web UI.", editable=True),
                 field("ws_port",  "WebSocket port",      "port",
-                      "TCP port for the plain WebSocket server."),
+                      "TCP port for the plain WebSocket server.", editable=True),
                 field("wss_port", "Secure WebSocket port", "port",
-                      "TCP port for WSS (TLS WebSocket). Leave empty to disable."),
+                      "TCP port for WSS (TLS WebSocket). Leave empty to disable.", editable=True),
             ],
         },
         {
             "id": "tls",
             "title": "TLS / WebSocket Security",
             "description": "Certificate paths used when wss_port is set.",
+            "section_mode": "form",
+            "api_section": None,
             "fields": [
                 field("cert_path", "Certificate directory", "path",
                       "Directory containing the TLS certificate and key files."),
@@ -232,73 +377,103 @@ def get_settings_sections(config: dict) -> list:
             "id": "monitoring",
             "title": "Monitoring",
             "description": "Heartbeat timing and alert re-notification behaviour.",
+            "section_mode": "form",
+            "api_section": "server",
             "fields": [
                 field("interval",  "Heartbeat interval", "duration",
-                      "Expected time between heartbeat messages from each client."),
+                      "Expected time between heartbeat messages from each client.", editable=True),
-                field("grace",     "Grace multiplier",   "number",
+                field("grace",     "Grace period",        "number",
-                      "A host is marked overdue after interval × grace seconds of silence."),
+                      "Extra seconds to wait after a missed heartbeat before sending notifications.", editable=True),
                 field("threshold_renotify_interval", "Re-notify interval", "duration",
-                      "How often to re-send notifications for ongoing threshold alerts."),
+                      "How often to re-send notifications for ongoing threshold alerts.", editable=True),
+                field("flap_count", "Flap count", "number",
+                      "Warning/critical notifications within the flap interval that mark a "
+                      "service as flapping and silence it. 0 disables flap detection.", editable=True),
+                field("flap_interval", "Flap interval", "number",
+                      "Minutes: the window flap count is measured over, and how long a service "
+                      "must stay OK before flapping ends.", editable=True),
                 field("autosave_interval", "Autosave interval", "duration",
                       "How often the server saves its state to disk."),
+                field("base_url", "Base URL", "text",
+                      "Base URL for notification links.", editable=True),
             ],
         },
         {
             "id": "persistence",
             "title": "Persistence & Logging",
             "description": "State file and event log settings.",
+            "section_mode": "form",
+            "api_section": "server",
             "fields": [
                 field("pickfile", "State file",   "path",
-                      "Path to the pickle file used to persist host state across restarts."),
+                      "Path to the pickle file used to persist host state across restarts.", editable=True),
                 field("logfile",  "Event log",    "path",
-                      "Path to the event log file."),
+                      "Path to the event log file.", editable=True),
             ],
         },
         {
             "id": "journal",
             "title": "Message Journal",
             "description": "All received heartbeat and plugin messages are journalled here.",
+            "section_mode": "form",
+            "api_section": "server",
             "fields": [
                 field("journal_enabled",     "Enabled",          "boolean",
-                      "Turn journalling on or off."),
+                      "Turn journalling on or off.", editable=True),
                 field("journal_dir",         "Journal directory","path",
-                      "Directory where journal files are written."),
+                      "Directory where journal files are written.", editable=True),
                 field("journal_file",        "Journal filename", "text",
                       "Base filename for the journal (rotated copies get a numeric suffix)."),
                 field("journal_max_size",    "Max file size",    "size",
-                      "Rotate the journal when it exceeds this size."),
+                      "Rotate the journal when it exceeds this size.", editable=True),
                 field("journal_max_backups", "Backup count",     "number",
-                      "Number of rotated journal files to keep."),
+                      "Number of rotated journal files to keep.", editable=True),
             ],
         },
         {
             "id": "dns",
             "title": "Dynamic DNS",
-            "description": "nsupdate-based DNS registration for dynamic hosts.",
+            "description": "nsupdate-based DNS registration via nsupdate(8).",
+            "section_mode": "form",
+            "api_section": "dns",
             "fields": [
                 field("nsupdate_bin", "nsupdate binary", "path",
-                      "Full path to the nsupdate executable."),
+                      "Path to the nsupdate binary.", editable=True),
-                field("dyndomains",   "Dynamic domains", "list",
+                field("rndc_key", "RNDC key file", "path",
-                      "DNS zones managed by nsupdate for dynamic hosts."),
+                      "Path to the rndc key file used to authenticate DNS updates.", editable=True),
-                field("drophosts",    "Drop hosts",      "list",
+                field("dyndomains", "Dynamic domains", "list",
-                      "Hostnames to silently ignore — no state, no alerts."),
+                      "Domains updated via nsupdate when a host with dyndns: true reports in.",
+                      editable=True),
             ],
         },
         {
             "id": "users",
             "title": "Users",
             "description": "Accounts defined in the config file. Password hashes are never shown.",
+            "section_mode": "form",
+            "api_section": "users",
             "users": users_list,
             "fields": [
                 field("default_owner", "Default owner", "text",
                       "Username that owns hosts with no explicit owner. "
-                      "Falls back to the first admin user."),
+                      "Falls back to the first admin user.", editable=True),
             ],
         },
+        {
+            "id": "oauth",
+            "title": "OAuth Providers",
+            "description": "OAuth2 login providers. Client secrets are masked.",
+            "section_mode": "form",
+            "api_section": "oauth",
+            "providers": oauth_providers,
+            "fields": [],
+        },
         {
             "id": "channels",
             "title": "Notification Channels",
             "description": "Named notification providers. Credentials are masked.",
+            "section_mode": "channels",
+            "api_section": "notification_channels",
             "channels": notif_channels,
             "fields": [
                 field("default_notification_channels", "Default channels", "list",
@@ -309,13 +484,29 @@ def get_settings_sections(config: dict) -> list:
             "id": "hosts",
             "title": "Hosts",
             "description": "Host definitions loaded from the config file.",
+            "section_mode": "hosts",
+            "api_section": "hosts",
             "hosts": hosts_list,
             "fields": [],
         },
+        {
+            "id": "thresholds",
+            "title": "Threshold Configurations",
+            "description": "Named alert threshold sets. Each defines warning/critical levels per metric.",
+            "section_mode": "thresholds",
+            "api_section": "thresholds",
+            "threshold_configs": threshold_config_list,
+            "fields": [
+                field("default_threshold_config", "Default config", "text",
+                      "Threshold config used for hosts with no explicit mapping.", editable=True),
+            ],
+        },
         {
             "id": "runtime",
             "title": "Runtime",
             "description": "Flags set at startup (require restart to change).",
+            "section_mode": "form",
+            "api_section": None,
             "fields": [
                 field("foreground", "Foreground mode", "boolean",
                       "Run in the foreground instead of daemonising."),
@@ -326,3 +517,27 @@ def get_settings_sections(config: dict) -> list:
             ],
         },
     ]
+
+    if not is_admin:
+        sections = [s for s in sections if s["id"] in ("channels", "hosts", "thresholds")]
+        for s in sections:
+            s["fields"] = []
+    return sections
+
+
+def get_settings_data(config: dict, threshold_checker=None, user=None) -> dict:
+    """Return sections list + auxiliary data for the settings template."""
+    sections = get_settings_sections(config, threshold_checker=threshold_checker, user=user)
+    is_admin = user is None or getattr(user, "admin", False)
+    username: str = getattr(user, "username", "") or ""
+    channels = config.get("notification_channels") or {}
+    threshold_cfgs = config.get("threshold_configs") or {}
+    if not is_admin:
+        channels = config_access.user_channels(channels, username)
+        threshold_cfgs = config_access.user_threshold_configs(threshold_cfgs, username)
+    return {
+        "sections": sections,
+        "all_channel_names": sorted(channels.keys()),
+        "all_usernames": sorted((config.get("users") or {}).keys()),
+        "all_threshold_configs": sorted(threshold_cfgs.keys()),
+    }

hbd/server/static/hbd-ui.css

@@ -0,0 +1,155 @@
+/* hbd-ui.css — shared design system for hbd pages.
+ *
+ * Green-biased neutrals, teal accent, mono identifiers. Pages link this
+ * after static/style.css and build on the tokens + components below.
+ * Introduced with the settings-page redesign; see settings.html for the
+ * fullest use of the row/editor idiom.
+ */
+
+:root {
+  --st-paper: #f4f6f4;
+  --st-surface: #ffffff;
+  --st-surface-2: #eef1ee;
+  --st-ink: #1d2420;
+  --st-muted: #5f6a63;
+  --st-faint: #8a948d;
+  --st-line: #dde3de;
+  --st-line-soft: #e8ece8;
+  --st-accent: #0e7280;
+  --st-accent-ink: #ffffff;
+  --st-accent-soft: #e0eef0;
+  --st-ok: #3f9c5a;
+  --st-ok-soft: #e2f2e7;
+  --st-warn: #d9982a;
+  --st-warn-soft: #f9efdc;
+  --st-crit: #d64545;
+  --st-crit-soft: #f9e3e3;
+  --st-badge-owner: #ede7f6; --st-badge-owner-ink: #5e35b1;
+  --st-badge-mgr: #e3f2fd;   --st-badge-mgr-ink: #1565c0;
+  --st-mono: ui-monospace, "SF Mono", "Cascadia Mono", Menlo, Consolas, monospace;
+}
+html[data-theme="dark"] {
+  --st-paper: #131816;
+  --st-surface: #1b211e;
+  --st-surface-2: #222925;
+  --st-ink: #e4eae6;
+  --st-muted: #9aa69e;
+  --st-faint: #6e7a72;
+  --st-line: #2c342f;
+  --st-line-soft: #252c28;
+  --st-accent: #4cc4d4;
+  --st-accent-ink: #0d2326;
+  --st-accent-soft: #143b40;
+  --st-ok: #5cba77;
+  --st-ok-soft: #1c2f22;
+  --st-warn: #e0ad4e;
+  --st-warn-soft: #33290f;
+  --st-crit: #e06c6c;
+  --st-crit-soft: #3a1a1a;
+  --st-badge-owner: #33294a; --st-badge-owner-ink: #b39ddb;
+  --st-badge-mgr: #1b3350;   --st-badge-mgr-ink: #90caf9;
+}
+
+/* ── sticky toolbar under the site nav ── */
+.st-toolbar {
+  position: sticky; top: var(--nav-h, 48px); z-index: 90;
+  display: flex; align-items: center; gap: 12px; flex-wrap: wrap;
+  padding: 9px 20px;
+  background: var(--st-surface); border-bottom: 1px solid var(--st-line);
+}
+.st-toolbar .brand { font-family: var(--st-mono); font-size: 15px; font-weight: 700; letter-spacing: -.02em; color: var(--st-ink); }
+.st-toolbar .brand .tld { color: var(--st-accent); }
+.st-toolbar .hintline { font-size: 12px; color: var(--st-faint); }
+.st-toolbar .spacer { flex: 1; }
+
+/* ── ECG pulse divider ── */
+.pulse { display: block; width: 100%; height: 14px; }
+.pulse polyline { fill: none; stroke: var(--st-accent); stroke-width: 1.5; opacity: .55; }
+
+/* ── buttons ── */
+.btn {
+  font: 600 13px/1 inherit; border: 1px solid var(--st-line); border-radius: 6px;
+  background: var(--st-surface); color: var(--st-ink); padding: 7px 14px; cursor: pointer;
+}
+.btn:focus-visible, a:focus-visible { outline: 2px solid var(--st-accent); outline-offset: 2px; }
+.btn.primary { background: var(--st-accent); border-color: var(--st-accent); color: var(--st-accent-ink); }
+.btn.quiet { border-color: transparent; background: transparent; color: var(--st-muted); }
+.btn.small { padding: 4px 10px; font-size: 12px; }
+.btn.danger { color: var(--st-crit); }
+.btn:disabled { opacity: .6; cursor: default; }
+
+/* ── section headers as config keys ── */
+section.st { margin-top: 26px; scroll-margin-top: calc(var(--bars-h, 94px) + 12px); }
+.sec-head { display: flex; align-items: baseline; gap: 10px; margin-bottom: 8px; flex-wrap: wrap; }
+.sec-head h2 { margin: 0; font-family: var(--st-mono); font-size: 15px; font-weight: 700; color: var(--st-ink); }
+.sec-head h2 .colon { color: var(--st-accent); }
+.sec-head .count { font-family: var(--st-mono); font-size: 11.5px; color: var(--st-faint); }
+.sec-head .sub { font-size: 12px; color: var(--st-muted); }
+.sec-head .add { margin-left: auto; }
+
+/* ── record list + row ── */
+.list { background: var(--st-surface); border: 1px solid var(--st-line); border-radius: 8px; overflow: hidden; }
+.row {
+  display: grid; grid-template-columns: 190px minmax(0,1fr) auto;
+  gap: 4px 14px; align-items: center;
+  padding: 8px 14px; border-bottom: 1px solid var(--st-line-soft);
+  transition: background .12s ease;
+}
+@media (prefers-reduced-motion: reduce) { .row { transition: none; } }
+.row:last-child { border-bottom: none; }
+.row:hover { background: var(--st-surface-2); }
+.row .id { font-family: var(--st-mono); font-size: 13px; font-weight: 600; color: var(--st-ink); display: flex; align-items: center; gap: 8px; min-width: 0; }
+.row .id .name { overflow: hidden; text-overflow: ellipsis; white-space: nowrap; }
+.row .act { display: flex; gap: 4px; justify-content: flex-end; }
+
+.dot { width: 8px; height: 8px; border-radius: 50%; flex: none; }
+.dot.ok { background: var(--st-ok); }
+.dot.warn { background: var(--st-warn); }
+.dot.crit { background: var(--st-crit); }
+.dot.off { background: transparent; border: 1.5px solid var(--st-faint); }
+
+.facts { display: flex; flex-wrap: wrap; gap: 4px 6px; align-items: center; min-width: 0; }
+.chip {
+  display: inline-flex; align-items: center; gap: 4px;
+  font-family: var(--st-mono); font-size: 11px; line-height: 1.3;
+  padding: 3px 7px; border-radius: 4px;
+  background: var(--st-surface-2); color: var(--st-muted);
+}
+.chip.owner { background: var(--st-badge-owner); color: var(--st-badge-owner-ink); }
+.chip.mgr { background: var(--st-badge-mgr); color: var(--st-badge-mgr-ink); }
+.chip.level { background: transparent; border: 1px solid var(--st-line); }
+.chip.private { background: transparent; border: 1px dashed var(--st-badge-owner-ink); color: var(--st-badge-owner-ink); }
+.chip.ok { background: var(--st-ok-soft); color: var(--st-ok); }
+.chip.warn { background: var(--st-warn-soft); color: var(--st-warn); }
+.chip.crit { background: var(--st-crit-soft); color: var(--st-crit); }
+.chip .k { color: var(--st-faint); }
+.chip .w { color: var(--st-warn); font-weight: 700; }
+.chip .c { color: var(--st-crit); font-weight: 700; }
+
+/* ── key/value rows (server groups, about page) ── */
+.kv .row { grid-template-columns: 230px minmax(0,1fr); padding: 6px 14px; }
+.group {
+  display: flex; align-items: center; gap: 10px;
+  padding: 7px 14px 5px; font-family: var(--st-mono); font-size: 11px; color: var(--st-faint);
+  background: var(--st-surface); border-bottom: 1px solid var(--st-line-soft);
+}
+.group .gbtn { margin-left: auto; }
+.kv .val { font-family: var(--st-mono); font-size: 12.5px; color: var(--st-muted); min-width: 0; }
+.kv .val a { color: var(--st-accent); text-decoration: none; }
+.kv .val a:hover { text-decoration: underline; }
+.kv .desc { display: block; font-size: 11px; color: var(--st-faint); margin-top: 2px; }
+.val-masked { color: var(--st-faint); letter-spacing: 2px; }
+
+.note { margin: 8px 2px 0; font-size: 12px; color: var(--st-faint); }
+
+/* ── handheld defaults for rows ── */
+@media (max-width: 760px) {
+  .st-toolbar { padding: 8px 10px; }
+  .st-toolbar .hintline { display: none; }
+  .row { grid-template-columns: minmax(0,1fr) auto; padding: 9px 12px; }
+  .row .id { grid-row: 1; grid-column: 1; }
+  .row .act { grid-row: 1; grid-column: 2; }
+  .row .facts { grid-column: 1 / -1; }
+  .kv .row { grid-template-columns: 1fr; }
+  section.st { scroll-margin-top: calc(var(--bars-h, 94px) + 50px); }
+}

hbd/server/static/hbd-ui.js

@@ -0,0 +1,16 @@
+/* hbd-ui.js — shared helpers for pages using hbd-ui.css.
+ *
+ * Keeps sticky toolbars stuck just below the fixed site nav, whose height
+ * varies with viewport width and wrapping.
+ */
+
+function _setStickyOffsets() {
+  const nav = document.querySelector('.nav');
+  const tb = document.querySelector('.st-toolbar');
+  const navH = nav ? nav.offsetHeight : 48;
+  const barsH = navH + (tb ? tb.offsetHeight : 46);
+  document.documentElement.style.setProperty('--nav-h', navH + 'px');
+  document.documentElement.style.setProperty('--bars-h', barsH + 'px');
+}
+window.addEventListener('resize', _setStickyOffsets);
+document.addEventListener('DOMContentLoaded', _setStickyOffsets);

hbd/server/static/style.css

@@ -185,7 +185,7 @@
   /* Slightly larger tap targets in tables */
   #ntable td, #ntable th {
     padding: 4px 6px !important;
-    font-size: 0.82em !important;
+    font-size: 1.00em !important;
   }
 
   /* Cards on plugin/alerts pages */

hbd/server/templates/about.html

@@ -1,176 +1,110 @@
 <!DOCTYPE html>
 <html>
   {% include 'head.html' %}
+  <link rel="stylesheet" href="/static/hbd-ui.css">
+  <script src="/static/hbd-ui.js"></script>
 
   <style>
-    html, body { overflow: visible; }
+    html, body { height: auto; overflow: visible; }
+    body { background: var(--st-paper); padding: 60px 0 0; }
 
-    .container {
+    .about-main { max-width: 700px; margin: 0 auto; padding: 0 20px 60px; }
-      max-width: 700px;
+
-      margin: 0 auto;
+    .hero {
+      display: flex; align-items: baseline; gap: 14px; flex-wrap: wrap;
+      padding: 26px 2px 0;
     }
-
+    .hero .hb-logo {
-    h1 {
+      font-family: var(--st-mono); font-size: 30px; font-weight: 700;
-      color: #333;
+      letter-spacing: -.03em; color: var(--st-ink);
-      margin-bottom: 4px;
-      font-size: 1.5em;
     }
-
+    .hero .hb-logo .tld { color: var(--st-accent); }
-    .subtitle {
+    .hero .version { font-family: var(--st-mono); font-size: 13px; color: var(--st-accent);
-      color: #666;
+      background: var(--st-accent-soft); border-radius: 999px; padding: 3px 12px; }
-      margin-bottom: 24px;
+    .hero .tagline { flex-basis: 100%; font-size: 13px; color: var(--st-muted); }
-      font-size: 0.9em;
-    }
-
-    .section {
-      background: #fff;
-      border-radius: 8px;
-      box-shadow: 0 1px 6px rgba(0,0,0,0.1);
-      padding: 20px 24px;
-      margin-bottom: 20px;
-    }
-
-    .section h2 {
-      font-size: 1em;
-      font-weight: 700;
-      color: #333;
-      margin: 0 0 16px;
-      padding-bottom: 10px;
-      border-bottom: 1px solid #eee;
-      text-transform: uppercase;
-      letter-spacing: 0.5px;
-    }
-
-    .info-row {
-      display: flex;
-      align-items: baseline;
-      padding: 8px 0;
-      border-bottom: 1px solid #f5f5f5;
-      font-size: 0.9em;
-    }
-    .info-row:last-child { border-bottom: none; }
-
-    .info-label {
-      width: 160px;
-      flex-shrink: 0;
-      color: #666;
-      font-size: 0.88em;
-    }
-
-    .info-value {
-      color: #222;
-      word-break: break-all;
-    }
-
-    .info-value a {
-      color: #0066cc;
-      text-decoration: none;
-    }
-    .info-value a:hover { text-decoration: underline; }
-
-    .version-badge {
-      display: inline-block;
-      padding: 3px 12px;
-      background: #e8f0fe;
-      color: #1a73e8;
-      border-radius: 12px;
-      font-size: 0.85em;
-      font-weight: 600;
-      font-family: monospace;
-    }
-
-    .hb-logo {
-      font-size: 2.5em;
-      font-weight: 700;
-      color: #0066cc;
-      letter-spacing: -1px;
-      margin-bottom: 6px;
-    }
-
-    .hb-tagline {
-      color: #555;
-      font-size: 0.95em;
-    }
-
-    .logo-section {
-      display: flex;
-      align-items: center;
-      gap: 20px;
-      padding: 8px 0 4px;
-    }
-
-    .logo-text { flex: 1; }
   </style>
 
   <body>
     {% include 'nav.html' %}
 
-    <div class="container">
+    <div class="st-toolbar">
-      <h1>{{ header }}</h1>
+      <span class="brand">hbd<span class="tld">·about</span></span>
-      <p class="subtitle">Heartbeat monitoring system</p>
+      <span class="hintline">heartbeat monitoring system</span>
+    </div>
+    <svg class="pulse" viewBox="0 0 1200 14" preserveAspectRatio="none" aria-hidden="true">
+      <polyline points="0,10 340,10 352,10 358,3 364,13 370,1 378,12 384,10 560,10 572,10 578,3 584,13 590,1 598,12 604,10 1200,10"/>
+    </svg>
 
-      <div class="section">
+    <div class="about-main">
-        <div class="logo-section">
+      <div class="hero">
-          <div class="logo-text">
+        <span class="hb-logo">heart<span class="tld">beat</span></span>
-            <div class="hb-logo">Heartbeat</div>
+        <span class="version">v{{ hbd_version }}</span>
-            <div class="hb-tagline">Lightweight host monitoring over UDP</div>
+        <span class="tagline">Lightweight host monitoring over UDP</span>
+      </div>
+
+      <section class="st">
+        <div class="sec-head">
+          <h2>version<span class="colon">:</span></h2>
+        </div>
+        <div class="list kv">
+          <div class="row">
+            <span class="id"><span class="name">server</span></span>
+            <span class="val">{{ hbd_version }}</span>
+          </div>
+          <div class="row">
+            <span class="id"><span class="name">python</span></span>
+            <span class="val">{{ python_version }}</span>
+          </div>
+          <div class="row">
+            <span class="id"><span class="name">license</span></span>
+            <span class="val">MIT</span>
           </div>
-          <span class="version-badge">v{{ hbd_version }}</span>
         </div>
-      </div>
+      </section>
 
-      <div class="section">
+      <section class="st">
-        <h2>Version</h2>
+        <div class="sec-head">
-        <div class="info-row">
+          <h2>runtime<span class="colon">:</span></h2>
-          <span class="info-label">Server version</span>
-          <span class="info-value">{{ hbd_version }}</span>
         </div>
-        <div class="info-row">
+        <div class="list kv">
-          <span class="info-label">Python</span>
+          <div class="row">
-          <span class="info-value">{{ python_version }}</span>
+            <span class="id"><span class="name">host</span></span>
+            <span class="val">{{ server_hostname }}</span>
+          </div>
+          <div class="row">
+            <span class="id"><span class="name">started</span></span>
+            <span class="val">{{ start_time_str }}</span>
+          </div>
+          <div class="row">
+            <span class="id"><span class="name">uptime</span></span>
+            <span class="val" id="uptime-value">{{ uptime_str }}</span>
+          </div>
+          <div class="row">
+            <span class="id"><span class="name">hosts_monitored</span></span>
+            <span class="val">{{ host_count }}</span>
+          </div>
         </div>
-        <div class="info-row">
+      </section>
-          <span class="info-label">License</span>
-          <span class="info-value">MIT</span>
-        </div>
-      </div>
 
-      <div class="section">
+      <section class="st">
-        <h2>Runtime</h2>
+        <div class="sec-head">
-        <div class="info-row">
+          <h2>contact<span class="colon">:</span></h2>
-          <span class="info-label">Host</span>
-          <span class="info-value">{{ server_hostname }}</span>
         </div>
-        <div class="info-row">
+        <div class="list kv">
-          <span class="info-label">Started</span>
+          <div class="row">
-          <span class="info-value">{{ start_time_str }}</span>
+            <span class="id"><span class="name">author</span></span>
+            <span class="val">Andreas Wrede</span>
+          </div>
+          <div class="row">
+            <span class="id"><span class="name">email</span></span>
+            <span class="val"><a href="mailto:aew.hbd@wrede.ca">aew.hbd@wrede.ca</a></span>
+          </div>
+          <div class="row">
+            <span class="id"><span class="name">repository</span></span>
+            <span class="val"><a href="https://git.wrede.ca/andreas/heartbeat" target="_blank" rel="noopener">git.wrede.ca/andreas/heartbeat</a></span>
+          </div>
         </div>
-        <div class="info-row">
+      </section>
-          <span class="info-label">Uptime</span>
-          <span class="info-value" id="uptime-value">{{ uptime_str }}</span>
-        </div>
-        <div class="info-row">
-          <span class="info-label">Hosts monitored</span>
-          <span class="info-value">{{ host_count }}</span>
-        </div>
-      </div>
-
-      <div class="section">
-        <h2>Contact &amp; Source</h2>
-        <div class="info-row">
-          <span class="info-label">Author</span>
-          <span class="info-value">Andreas Wrede</span>
-        </div>
-        <div class="info-row">
-          <span class="info-label">Email</span>
-          <span class="info-value"><a href="mailto:aew@wrede.ca">aew@wrede.ca</a></span>
-        </div>
-        <div class="info-row">
-          <span class="info-label">Repository</span>
-          <span class="info-value"><a href="https://git.wrede.ca/andreas/heartbeat" target="_blank" rel="noopener">git.wrede.ca/andreas/heartbeat</a></span>
-        </div>
-      </div>
-
     </div>
 
     <script>

hbd/server/templates/alerts.html

@@ -1,523 +1,236 @@
 <!DOCTYPE html>
 <html>
   {% include 'head.html' %}
+  <link rel="stylesheet" href="/static/hbd-ui.css">
+  <script src="/static/hbd-ui.js"></script>
 
   <style>
+    html, body { height: auto; overflow: visible; }
+    body { background: var(--st-paper); padding: 60px 0 0; }
 
-    .container {
+    .alerts-main { max-width: 1100px; margin: 0 auto; padding: 0 20px 60px; }
-      max-width: 1400px;
+
-      margin: 0 auto;
+    /* summary tiles */
+    .stats { display: flex; gap: 10px; flex-wrap: wrap; padding-top: 18px; }
+    .stat {
+      flex: 1 1 140px; display: flex; align-items: baseline; gap: 10px;
+      background: var(--st-surface); border: 1px solid var(--st-line); border-radius: 8px;
+      padding: 10px 14px;
     }
+    .stat .num { font-family: var(--st-mono); font-size: 24px; font-weight: 700; font-variant-numeric: tabular-nums; }
+    .stat .lbl { font-family: var(--st-mono); font-size: 11.5px; color: var(--st-faint); }
+    .stat.crit .num { color: var(--st-crit); }
+    .stat.warn .num { color: var(--st-warn); }
+    .stat.ok .num { color: var(--st-ok); }
 
-    h1 { color: #333; margin-bottom: 5px; margin-top: 15px; font-size: 1.5em; }
+    /* filter bar */
-
+    .filterbar { display: flex; gap: 6px; align-items: center; flex-wrap: wrap; margin: 14px 0 0; }
-    .subtitle {
+    .filterbar .fbtn {
-      color: #666;
+      font-family: var(--st-mono); font-size: 12px; cursor: pointer;
-      margin-bottom: 30px;
+      color: var(--st-muted); background: var(--st-surface); border: 1px solid var(--st-line);
+      border-radius: 999px; padding: 5px 12px;
     }
-
+    .filterbar .fbtn.active { background: var(--st-accent-soft); border-color: var(--st-accent); color: var(--st-accent); font-weight: 600; }
-    .summary-cards {
+    .filterbar input {
-      display: flex;
+      font: 12px var(--st-mono); color: var(--st-ink);
-      flex-wrap: wrap;
+      background: var(--st-surface); border: 1px solid var(--st-line); border-radius: 999px;
-      gap: 10px;
+      padding: 5px 12px; min-width: 170px;
-      margin-bottom: 16px;
     }
+    .filterbar input.invalid { border-color: var(--st-crit); }
+    .filterbar .upd { margin-left: auto; font-size: 11.5px; color: var(--st-faint); }
 
-    .summary-card {
+    /* alert row specifics */
-      background: white;
+    .row.alert-crit { border-left: 3px solid var(--st-crit); padding-left: 11px; }
-      border-radius: 6px;
+    .row.alert-warn { border-left: 3px solid var(--st-warn); padding-left: 11px; }
-      padding: 6px 14px;
+    .row.acked { opacity: .55; border-left-style: dashed; }
-      box-shadow: 0 1px 4px rgba(0,0,0,0.1);
+    .row .id a { color: inherit; text-decoration: none; }
-      display: flex;
+    .row .id a:hover { color: var(--st-accent); }
-      align-items: center;
+    .acked-chip { color: var(--st-ok); font-size: 12px; white-space: nowrap; }
-      gap: 8px;
+
-      border-left: 4px solid #ddd;
+    .empty {
+      padding: 36px 14px; text-align: center; color: var(--st-faint); font-size: 13px;
     }
-
+    .empty .big { font-size: 26px; color: var(--st-ok); display: block; margin-bottom: 6px; }
-    .summary-card.critical { border-left-color: #ea1e0f; }
+    .errorbox {
-    .summary-card.warning  { border-left-color: #ff9800; }
+      background: var(--st-crit-soft); color: var(--st-crit);
-    .summary-card.ok       { border-left-color: #4caf50; }
+      border-radius: 8px; padding: 14px 16px; font-size: 13px;
-
-    .summary-number {
-      font-size: 1.4em;
-      font-weight: bold;
-      line-height: 1;
-    }
-
-    .summary-number.critical { color: #ea1e0f; }
-    .summary-number.warning  { color: #ff9800; }
-    .summary-number.ok       { color: #4caf50; }
-
-    .summary-label {
-      color: #666;
-      font-size: 0.85em;
-    }
-
-    .filters {
-      background: white;
-      border-radius: 8px;
-      padding: 15px;
-      margin-bottom: 20px;
-      box-shadow: 0 2px 8px rgba(0,0,0,0.1);
-      display: flex;
-      gap: 15px;
-      align-items: center;
-    }
-
-    .filter-label {
-      font-weight: bold;
-      color: #555;
-    }
-
-    .filter-button {
-      padding: 8px 16px;
-      border: 2px solid #ddd;
-      background: white;
-      border-radius: 20px;
-      cursor: pointer;
-      transition: all 0.2s;
-      font-size: 0.9em;
-    }
-
-    .filter-button:hover {
-      border-color: #2196f3;
-    }
-
-    .filter-button.active {
-      background: #2196f3;
-      color: white;
-      border-color: #2196f3;
-    }
-
-    .alerts-container {
-      background: white;
-      border-radius: 8px;
-      padding: 20px;
-      box-shadow: 0 2px 8px rgba(0,0,0,0.1);
-    }
-
-    .alert-item {
-      border-left: 5px solid #ddd;
-      padding: 15px;
-      margin-bottom: 15px;
-      background: #fafafa;
-      border-radius: 4px;
-      display: flex;
-      justify-content: space-between;
-      align-items: center;
-      transition: all 0.2s;
-    }
-    
-    .alert-item.acknowledged {
-      opacity: 0.8;
-      background: #f0f0f0;
-    }
-
-    .alert-item:hover {
-      box-shadow: 0 2px 8px rgba(0,0,0,0.1);
-      transform: translateX(5px);
-    }
-
-    .alert-item.critical {
-      border-left-color: #f44336;
-      background: #ffebee;
-    }
-
-    .alert-item.warning {
-      border-left-color: #ff9800;
-      background: #fff3e0;
-    }
-
-    .alert-item.unknown {
-      border-left-color: #9e9e9e;
-      background: #f5f5f5;
-    }
-
-    .alert-main {
-      flex: 1;
-    }
-
-    .alert-header {
-      display: flex;
-      align-items: center;
-      gap: 15px;
-      margin-bottom: 8px;
-    }
-
-    .alert-level {
-      padding: 4px 12px;
-      border-radius: 12px;
-      font-size: 0.75em;
-      font-weight: bold;
-      text-transform: uppercase;
-      letter-spacing: 0.5px;
-    }
-
-    .alert-level.critical {
-      background: #f44336;
-      color: white;
-    }
-
-    .alert-level.warning {
-      background: #ff9800;
-      color: white;
-    }
-
-    .alert-level.unknown {
-      background: #9e9e9e;
-      color: white;
-    }
-
-    .alert-hostname {
-      font-weight: bold;
-      color: #333;
-      font-size: 1.1em;
-    }
-
-    .alert-metric {
-      color: #666;
-      font-family: 'Courier New', monospace;
-      font-size: 0.9em;
-    }
-
-    .alert-details {
-      display: flex;
-      gap: 20px;
-      color: #666;
-      font-size: 0.9em;
-    }
-
-    .alert-value {
-      font-weight: bold;
-      color: #333;
-    }
-
-    .alert-duration {
-      color: #999;
-      font-size: 0.85em;
-    }
-    
-    .alert-actions {
-      display: flex;
-      flex-direction: column;
-      gap: 8px;
-      margin-left: 15px;
-    }
-    
-    .acknowledge-btn {
-      padding: 8px 16px;
-      background: #2196f3;
-      color: white;
-      border: none;
-      border-radius: 4px;
-      cursor: pointer;
-      font-size: 0.85em;
-      transition: all 0.2s;
-      white-space: nowrap;
-    }
-    
-    .acknowledge-btn:hover {
-      background: #1976d2;
-      transform: scale(1.05);
-    }
-    
-    .acknowledge-btn:disabled {
-      background: #ccc;
-      cursor: not-allowed;
-      transform: none;
-    }
-    
-    .acknowledged-badge {
-      padding: 4px 8px;
-      background: #4caf50;
-      color: white;
-      border-radius: 4px;
-      font-size: 0.75em;
-      text-align: center;
-      white-space: nowrap;
-    }
-
-    .no-alerts {
-      text-align: center;
-      padding: 60px 20px;
-      color: #999;
-    }
-
-    .no-alerts-icon {
-      font-size: 4em;
-      margin-bottom: 20px;
-    }
-
-    .loading {
-      text-align: center;
-      padding: 40px;
-      color: #666;
-    }
-
-    .error {
-      background: #ffebee;
-      border-left: 4px solid #f44336;
-      padding: 20px;
-      margin: 20px 0;
-      border-radius: 4px;
-      color: #c62828;
-    }
-
-    .refresh-info {
-      text-align: center;
-      color: #999;
-      font-size: 0.85em;
-      margin-top: 20px;
-      padding-top: 20px;
-      border-top: 1px solid #e0e0e0;
-    }
-
-    .last-update {
-      color: #666;
-      font-size: 0.9em;
-      text-align: right;
-      margin-bottom: 15px;
     }
   </style>
 
   <body>
     {% include 'nav.html' %}
 
-    <div class="container">
+    <div class="st-toolbar">
-      <h1>{{ header }}</h1>
+      <span class="brand">hbd<span class="tld">·alerts</span></span>
-      <p class="subtitle">Real-time monitoring alerts and threshold violations</p>
+      <span class="hintline">threshold violations and reachability — refreshes every 15 s</span>
+    </div>
+    <svg class="pulse" viewBox="0 0 1200 14" preserveAspectRatio="none" aria-hidden="true">
+      <polyline points="0,10 340,10 352,10 358,3 364,13 370,1 378,12 384,10 560,10 572,10 578,3 584,13 590,1 598,12 604,10 1200,10"/>
+    </svg>
 
-      <div class="summary-cards" id="summary-cards">
+    <div class="alerts-main">
-        <div class="summary-card critical">
+      <div class="stats">
-          <div class="summary-label">Critical</div>
+        <div class="stat crit"><span class="num" id="critical-count">–</span><span class="lbl">critical</span></div>
-          <div class="summary-number critical" id="critical-count">-</div>
+        <div class="stat warn"><span class="num" id="warning-count">–</span><span class="lbl">warning</span></div>
-        </div>
+        <div class="stat ok"><span class="num" id="host-count">–</span><span class="lbl">hosts</span></div>
-        <div class="summary-card warning">
-          <div class="summary-label">Warning</div>
-          <div class="summary-number warning" id="warning-count">-</div>
-        </div>
-        <div class="summary-card ok">
-          <div class="summary-label">Total Hosts</div>
-          <div class="summary-number ok" id="host-count">-</div>
-        </div>
       </div>
 
-      <div class="filters">
+      <div class="filterbar">
-        <span class="filter-label">Show:</span>
+        <button class="fbtn active" onclick="filterAlerts('all', this)">all</button>
-        <button class="filter-button active" onclick="filterAlerts('all')">All</button>
+        <button class="fbtn" onclick="filterAlerts('critical', this)">critical</button>
-        <button class="filter-button" onclick="filterAlerts('critical')">Critical Only</button>
+        <button class="fbtn" onclick="filterAlerts('warning', this)">warning</button>
-        <button class="filter-button" onclick="filterAlerts('warning')">Warning Only</button>
+        <input id="host-filter" type="text" placeholder="host filter (regex)" oninput="onHostFilterInput(this)">
+        <span class="upd">updated <span id="last-update-time">never</span></span>
       </div>
 
-      <div class="alerts-container">
+      <section class="st" id="alerts">
-        <div class="last-update">Last updated: <span id="last-update-time">Never</span></div>
+        <div class="sec-head">
-        <div id="alerts-list">
+          <h2>alerts<span class="colon">:</span></h2>
-          <div class="loading">Loading alerts...</div>
+          <span class="count" id="alert-count"></span>
         </div>
-        <div class="refresh-info">
+        <div class="list" id="alerts-list">
-          Auto-refreshing every 15 seconds
+          <div class="empty">Loading alerts…</div>
         </div>
-      </div>
+      </section>
     </div>
 
     <script>
       let currentFilter = 'all';
       let allAlerts = [];
+      let hostFilterRe = null;
+
+      function escHtml(s) {
+        return String(s).replace(/&/g,'&amp;').replace(/</g,'&lt;').replace(/>/g,'&gt;').replace(/"/g,'&quot;').replace(/'/g,'&#39;');
+      }
 
       async function loadAlerts() {
         try {
           const response = await fetch('/api/0/alerts');
-          if (!response.ok) {
+          if (!response.ok) throw new Error(`HTTP ${response.status}`);
-            throw new Error(`HTTP ${response.status}`);
-          }
-          
           const data = await response.json();
           allAlerts = data.alerts;
 
-          // Update summary cards
           document.getElementById('critical-count').textContent = data.summary.critical || 0;
           document.getElementById('warning-count').textContent = data.summary.warning || 0;
           document.getElementById('host-count').textContent = data.host_count || 0;
-          
-          // Update last update time
           document.getElementById('last-update-time').textContent = new Date().toLocaleTimeString();
 
-          // Render alerts
           renderAlerts(allAlerts);
-          
         } catch (error) {
           document.getElementById('alerts-list').innerHTML =
-            `<div class="error">Failed to load alerts: ${error.message}</div>`;
+            `<div class="errorbox">Failed to load alerts: ${escHtml(error.message)}. Retrying automatically.</div>`;
         }
       }
 
       function renderAlerts(alerts) {
         const container = document.getElementById('alerts-list');
 
-        // Filter alerts based on current filter
+        let filtered = alerts;
-        let filteredAlerts = alerts;
         if (currentFilter !== 'all') {
-          filteredAlerts = alerts.filter(alert => 
+          filtered = filtered.filter(a => a.level.toLowerCase() === currentFilter);
-            alert.level.toLowerCase() === currentFilter
+        }
-          );
+        if (hostFilterRe) {
+          filtered = filtered.filter(a => hostFilterRe.test(a.hostname));
         }
 
-        if (filteredAlerts.length === 0) {
+        document.getElementById('alert-count').textContent =
-          if (currentFilter === 'all' && alerts.length === 0) {
+          filtered.length === alerts.length ? alerts.length : `${filtered.length} of ${alerts.length}`;
-            container.innerHTML = `
+
-              <div class="no-alerts">
+        if (filtered.length === 0) {
-                <div class="no-alerts-icon">✓</div>
+          container.innerHTML = (currentFilter === 'all' && !hostFilterRe && alerts.length === 0)
-                <h2>All Systems Normal</h2>
+            ? `<div class="empty"><span class="big">✓</span>All systems normal — no active alerts.</div>`
-                <p>No active alerts at this time</p>
+            : `<div class="empty">No matching alerts.</div>`;
-              </div>
-            `;
-          } else {
-            container.innerHTML = `
-              <div class="no-alerts">
-                <p>No ${currentFilter} alerts</p>
-              </div>
-            `;
-          }
           return;
         }
 
-        let html = '';
+        container.innerHTML = filtered.map(renderAlert).join('');
-        for (const alert of filteredAlerts) {
-          html += renderAlert(alert);
-        }
-        container.innerHTML = html;
       }
 
       function renderAlert(alert) {
         const level = alert.level.toLowerCase();
+        const cls = level === 'critical' ? 'crit' : (level === 'warning' ? 'warn' : 'level');
         const duration = getDuration(alert.since);
-        const acknowledged = alert.acknowledged || false;
+        const acked = alert.acknowledged || false;
 
-        // Use formatted message if available, otherwise build from individual fields
+        const metric = (alert.metric_path.includes('.')
-        let valueText = `Value: <span class="alert-value">${formatValue(alert.last_value)}</span>`;
+          ? alert.metric_path.slice(alert.metric_path.indexOf('.') + 1)
+          : alert.metric_path).replace(/_status_code$/, '');
+
+        let chips = `<span class="chip ${cls}">${escHtml(alert.level)}</span>`;
+        chips += `<span class="chip"><span class="k">metric</span> ${escHtml(metric)}</span>`;
         if (alert.formatted_message) {
-          valueText += ` <span class="threshold-info">${alert.formatted_message}</span>`;
+          chips += `<span class="chip">${escHtml(alert.formatted_message)}</span>`;
-        } else if (alert.threshold_value !== undefined && alert.threshold_value !== null && alert.operator) {
-          valueText += ` <span class="threshold-info">(threshold: ${alert.operator} ${formatValue(alert.threshold_value)})</span>`;
-        }
-        
-        // Build actions section
-        let actionsHtml = '';
-        if (acknowledged) {
-          actionsHtml = `
-            <div class="alert-actions">
-              <div class="acknowledged-badge">✓ Acknowledged</div>
-            </div>
-          `;
         } else {
-          actionsHtml = `
+          chips += `<span class="chip"><span class="k">value</span> ${escHtml(formatValue(alert.last_value))}</span>`;
-            <div class="alert-actions">
+          if (alert.threshold_value !== undefined && alert.threshold_value !== null && alert.operator) {
-              <button class="acknowledge-btn" onclick="acknowledgeAlert('${alert.hostname}', '${alert.metric_path}', event)">
+            chips += `<span class="chip level">${escHtml(alert.operator)} ${escHtml(formatValue(alert.threshold_value))}</span>`;
-                Acknowledge
+          }
-              </button>
-            </div>
-          `;
         }
+        if (alert.recovery_threshold !== undefined && alert.recovery_threshold !== null) {
+          const recOp = (alert.operator === '>' || alert.operator === '>=') ? '<' : '>';
+          chips += `<span class="chip level">recovers ${recOp} ${escHtml(formatValue(alert.recovery_threshold))}</span>`;
+        }
+        chips += `<span class="chip"><span class="k">for</span> ${duration}</span>`;
+
+        const act = acked
+          ? `<span class="acked-chip">✓ acknowledged</span>`
+          : `<button class="btn small" onclick="acknowledgeAlert('${escHtml(alert.hostname)}', '${escHtml(alert.metric_path)}', event)">Acknowledge</button>`;
 
         return `
-          <div class="alert-item ${level} ${acknowledged ? 'acknowledged' : ''}">
+          <div class="row alert-${cls === 'level' ? 'warn' : cls}${acked ? ' acked' : ''}">
-            <div class="alert-main">
+            <span class="id">
-              <div class="alert-header">
+              <span class="dot ${cls === 'crit' ? 'crit' : 'warn'}"></span>
-                <span class="alert-level ${level}">${alert.level}</span>
+              <a class="name" href="/plugins#${encodeURIComponent(alert.hostname)}">${escHtml(alert.hostname)}</a>
-                <span class="alert-hostname">${alert.hostname}</span>
+            </span>
-              </div>
+            <span class="facts">${chips}</span>
-              <div class="alert-metric">${alert.metric_path}</div>
+            <span class="act">${act}</span>
-              <div class="alert-details">
+          </div>`;
-                <span>${valueText}</span>
-                <span class="alert-duration">Active for ${duration}</span>
-              </div>
-            </div>
-            ${actionsHtml}
-          </div>
-        `;
       }
 
       function formatValue(value) {
         if (typeof value === 'number') {
-          if (value > 1000) {
+          if (value > 1000) return value.toLocaleString();
-            return value.toLocaleString();
-          }
           return value.toFixed(2);
         }
         return value;
       }
 
       function getDuration(timestamp) {
-        const now = Date.now() / 1000;
+        const seconds = Math.floor(Date.now() / 1000 - timestamp);
-        const seconds = Math.floor(now - timestamp);
+        if (seconds < 60) return `${seconds}s`;
-        
+        if (seconds < 3600) return `${Math.floor(seconds / 60)}m`;
-        if (seconds < 60) {
+        if (seconds < 86400) {
-          return `${seconds}s`;
+          return `${Math.floor(seconds / 3600)}h ${Math.floor((seconds % 3600) / 60)}m`;
-        } else if (seconds < 3600) {
-          return `${Math.floor(seconds / 60)}m`;
-        } else if (seconds < 86400) {
-          const hours = Math.floor(seconds / 3600);
-          const minutes = Math.floor((seconds % 3600) / 60);
-          return `${hours}h ${minutes}m`;
-        } else {
-          const days = Math.floor(seconds / 86400);
-          const hours = Math.floor((seconds % 86400) / 3600);
-          return `${days}d ${hours}h`;
         }
+        return `${Math.floor(seconds / 86400)}d ${Math.floor((seconds % 86400) / 3600)}h`;
       }
 
-      function filterAlerts(filter) {
+      function filterAlerts(filter, btn) {
         currentFilter = filter;
-        
+        document.querySelectorAll('.filterbar .fbtn').forEach(b => b.classList.remove('active'));
-        // Update active button
+        btn.classList.add('active');
-        document.querySelectorAll('.filter-button').forEach(btn => {
-          btn.classList.remove('active');
-        });
-        event.target.classList.add('active');
-        
-        // Re-render with new filter
         renderAlerts(allAlerts);
       }
 
       async function acknowledgeAlert(hostname, metricPath, event) {
-        // Prevent event bubbling
+        if (event) event.stopPropagation();
-        if (event) {
-          event.stopPropagation();
-        }
-        
-        // Disable the button
         const button = event.target;
         button.disabled = true;
-        button.textContent = 'Acknowledging...';
+        button.textContent = 'Acknowledging…';
-        
         try {
           const response = await fetch('/api/0/alerts/acknowledge', {
             method: 'POST',
-            headers: {
+            headers: {'Content-Type': 'application/json'},
-              'Content-Type': 'application/json',
+            body: JSON.stringify({hostname: hostname, metric_path: metricPath}),
-            },
-            body: JSON.stringify({
-              hostname: hostname,
-              metric_path: metricPath,
-            }),
           });
-          
+          if (!response.ok) throw new Error(`HTTP ${response.status}`);
-          if (!response.ok) {
-            throw new Error(`HTTP ${response.status}`);
-          }
-          
           const result = await response.json();
-          
+          const a = allAlerts.find(x => x.hostname === hostname && x.metric_path === metricPath);
-          // Update the alert in our local data
+          if (a) {
-          const alert = allAlerts.find(a => a.hostname === hostname && a.metric_path === metricPath);
+            a.acknowledged = true;
-          if (alert) {
+            a.acknowledged_at = result.acknowledged_at;
-            alert.acknowledged = true;
-            alert.acknowledged_at = result.acknowledged_at;
           }
-          
-          // Re-render alerts
           renderAlerts(allAlerts);
-          
         } catch (error) {
           alert(`Failed to acknowledge alert: ${error.message}`);
           button.disabled = false;
@@ -525,10 +238,34 @@
         }
       }
 
-      // Auto-refresh every 15 seconds
+      function onHostFilterInput(input) {
+        const val = input.value.trim();
+        if (!val) {
+          hostFilterRe = null;
+          input.classList.remove('invalid');
+        } else {
+          try {
+            hostFilterRe = new RegExp(val, 'i');
+            input.classList.remove('invalid');
+          } catch (_) {
+            hostFilterRe = null;
+            input.classList.add('invalid');
+          }
+        }
+        renderAlerts(allAlerts);
+      }
+
       setInterval(loadAlerts, 15000);
 
-      // Initial load
+      (function () {
+        const param = new URLSearchParams(window.location.search).get('filter');
+        if (param) {
+          const input = document.getElementById('host-filter');
+          input.value = param;
+          onHostFilterInput(input);
+        }
+      })();
+
       loadAlerts();
     </script>
   </body>

hbd/server/templates/foot.html

@@ -1,5 +1,5 @@
  <footer>
  <div id="copyright">
-    &copy;2002-2026 <A HREF="mailto:andreas@wrede.ca">Andreas Wrede</A> All Rights Reserved.</p>
+    &copy;2002-2026 <A HREF="mailto:aew.hbd@wrede.ca">Andreas Wrede</A> All Rights Reserved.</p>
     </div>
 </footer>

hbd/server/templates/head.html

@@ -5,7 +5,68 @@
     <link rel="icon" href="/static/images/favicon.ico" sizes="32x32" />
     <title>{{ title }}</title>
     {% if extra_scripts %}<script src="{{ extra_scripts }}"></script>{% endif %}
+    <script>
+    /* Apply saved theme before first paint to avoid flash */
+    (function() {
+      try {
+        var p = localStorage.getItem('hbd_theme') || 'auto';
+        var dark = p === 'dark' || (p === 'auto' && window.matchMedia('(prefers-color-scheme: dark)').matches);
+        if (dark) document.documentElement.setAttribute('data-theme', 'dark');
+      } catch(e) {}
+    })();
+    </script>
     <style>
+      /* ── Theme variables ── */
+      :root {
+        --bg:           #f5f5f5;
+        --surface:      #ffffff;
+        --surface-2:    #f8f8f8;
+        --surface-3:    #f5f5f5;
+        --text:         #222222;
+        --text-2:       #333333;
+        --text-3:       #555555;
+        --text-sec:     #666666;
+        --text-muted:   #888888;
+        --text-dim:     #aaaaaa;
+        --text-ghost:   #cccccc;
+        --border:       #e0e0e0;
+        --border-2:     #eeeeee;
+        --border-3:     #f0f0f0;
+        --border-4:     #f5f5f5;
+        --link:         #0066cc;
+        --nav-bg:       #ffffff;
+        --input-bg:     #ffffff;
+        --input-border: #cccccc;
+        --shadow-sm:    rgba(0,0,0,.08);
+        --shadow:       rgba(0,0,0,.10);
+        --shadow-nav:   rgba(0,0,0,.10);
+      }
+      html[data-theme="dark"] {
+        color-scheme: dark;
+        --bg:           #111827;
+        --surface:      #1f2937;
+        --surface-2:    #283447;
+        --surface-3:    #374151;
+        --text:         #e5e7eb;
+        --text-2:       #d1d5db;
+        --text-3:       #9ca3af;
+        --text-sec:     #9ca3af;
+        --text-muted:   #6b7280;
+        --text-dim:     #4b5563;
+        --text-ghost:   #374151;
+        --border:       #374151;
+        --border-2:     #2d3748;
+        --border-3:     #253040;
+        --border-4:     #1e2a38;
+        --link:         #60a5fa;
+        --nav-bg:       #1f2937;
+        --input-bg:     #283447;
+        --input-border: #4b5563;
+        --shadow-sm:    rgba(0,0,0,.30);
+        --shadow:       rgba(0,0,0,.40);
+        --shadow-nav:   rgba(0,0,0,.40);
+      }
+
       /* ── Reset / shared baseline ── */
       *, *::before, *::after { box-sizing: border-box; }
       html {
@@ -16,10 +77,11 @@
         margin: 0;
         padding: 10px;
         padding-top: 60px;
-        background: #f5f5f5;
+        background: var(--bg);
+        color: var(--text);
       }
-      h1 { font-size: 1.5em; color: #333; margin: 0 0 5px; }
+      h1 { font-size: 1.5em; color: var(--text-2); margin: 0 0 5px; }
-      h2 { font-size: 1.1em; color: #333; margin: 0 0 8px; }
+      h2 { font-size: 1.1em; color: var(--text-2); margin: 0 0 8px; }
       p  { margin: 0; }
 
       /* Navigation bar — shared across all pages */
@@ -29,9 +91,9 @@
         left: 0;
         right: 0;
         z-index: 200;
-        background: #fff;
+        background: var(--nav-bg);
         padding: 6px 12px;
-        box-shadow: 0 2px 4px rgba(0,0,0,.1);
+        box-shadow: 0 2px 4px var(--shadow-nav);
         display: flex;
         align-items: center;
         justify-content: space-between;
@@ -42,25 +104,25 @@
       .nav a {
         margin-right: 20px;
         text-decoration: none;
-        color: #0066cc;
+        color: var(--link);
         font-weight: 500;
         font-size: 0.9em;
       }
       .nav a:hover { text-decoration: underline; }
-      .nav a.active { color: #333; font-weight: bold; }
+      .nav a.active { color: var(--text-2); font-weight: bold; }
       .nav-user {
         display: flex;
         align-items: center;
         gap: 8px;
         text-decoration: none;
-        color: #333;
+        color: var(--text-2);
         font-size: 0.9em;
         font-weight: 500;
         padding: 4px 8px;
         border-radius: 20px;
         transition: background 0.15s;
       }
-      .nav-user:hover { background: #f0f4ff; text-decoration: none; }
+      .nav-user:hover { background: var(--surface-2); text-decoration: none; }
       .nav-username {
         max-width: 0;
         overflow: hidden;
@@ -81,7 +143,7 @@
       .nav-initials {
         width: 28px; height: 28px;
         border-radius: 50%;
-        background: #0066cc;
+        background: var(--link);
         color: #fff;
         display: flex;
         align-items: center;
@@ -106,7 +168,7 @@
       .nav-hamburger span {
         display: block;
         height: 3px;
-        background: #555;
+        background: var(--text-muted);
         border-radius: 2px;
       }
 
@@ -118,19 +180,51 @@
           flex-direction: column;
           align-items: flex-start;
           padding-top: 8px;
-          border-top: 1px solid #eee;
+          border-top: 1px solid var(--border-2);
           order: 3;
         }
         .nav-links.nav-open { display: flex; }
         .nav-links a { margin-right: 0; padding: 6px 0; font-size: 1em; }
       }
 
+      /* ── Global dark-mode: inputs ── */
+      html[data-theme="dark"] input:not([type=checkbox]):not([type=radio]),
+      html[data-theme="dark"] select,
+      html[data-theme="dark"] textarea {
+        background-color: var(--input-bg);
+        border-color: var(--input-border);
+        color: var(--text);
+      }
+
+      /* Pending config publish button */
+      .nav-publish-btn {
+        background: #e65100;
+        color: #fff;
+        border: none;
+        border-radius: 4px;
+        padding: 4px 10px;
+        font-size: 0.82em;
+        font-weight: 600;
+        cursor: pointer;
+        flex-shrink: 0;
+        white-space: nowrap;
+        margin-left: auto;
+      }
+      .nav-publish-btn:hover { background: #bf360c; }
+      .nav-publish-btn:disabled { opacity: 0.7; cursor: default; }
+
       /* Swiss railway clock — nav */
-      .nav-clock {
+      .nav-pie {
         flex-shrink: 0;
         line-height: 0;
         margin-left: auto;
         padding: 4px 4px 4px 0;
+      }
+      #alert-pie { display: block; cursor: default; }
+      .nav-clock {
+        flex-shrink: 0;
+        line-height: 0;
+        padding: 4px 4px 4px 0;
         cursor: pointer;
       }
       #swiss-clock { display: block; }
@@ -208,7 +302,7 @@
         ctx.restore();
       }
 
-      hand((m + s / 60) / 60 * Math.PI * 2 - Math.PI / 2,
+      hand((sFrac >= 58.5 ? m + 1 : m) / 60 * Math.PI * 2 - Math.PI / 2,
            R * 0.88, -R * 0.12, SIZE * 0.027, '#222');           /* minute */
       hand((h + m / 60) / 12 * Math.PI * 2 - Math.PI / 2,
            R * 0.58, -R * 0.12, SIZE * 0.039, '#222');           /* hour   */
@@ -256,6 +350,17 @@
       setTimeout(clockTick, delay);
     }
 
+    /* Keep auto-theme in sync with system setting changes */
+    try {
+      window.matchMedia('(prefers-color-scheme: dark)').addEventListener('change', function(e) {
+        var pref = localStorage.getItem('hbd_theme') || 'auto';
+        if (pref === 'auto') {
+          if (e.matches) { document.documentElement.setAttribute('data-theme', 'dark'); }
+          else { document.documentElement.removeAttribute('data-theme'); }
+        }
+      });
+    } catch(e) {}
+
     document.addEventListener('DOMContentLoaded', function() {
       /* Start the shared tick loop */
       clockTick();

hbd/server/templates/live.html

@@ -1,566 +1,408 @@
 <!DOCTYPE html>
 <html>
   {% include 'head.html' %}
+  <link rel="stylesheet" href="/static/hbd-ui.css">
+  <script src="/static/hbd-ui.js"></script>
 
   <style>
-    body {
+    html, body { height: auto; overflow: visible; }
-      display: flex;
+    body { background: var(--st-paper); padding: 60px 0 0; }
-      flex-direction: column;
-      height: 100vh;
-      overflow: hidden;
-    }
 
-    @media (max-width: 640px) {
+    .live-main { max-width: 1180px; margin: 0 auto; padding: 0 20px 70px; }
-      body {
-        height: auto;
-        min-height: 100vh;
-        overflow: auto;
-        flex-direction: column;
-      }
-      .container {
-        max-height: none;
-        overflow: visible;
-      }
-      .table-section {
-        max-height: 55vh;
-      }
-      .log-section {
-        flex: none;
-        max-height: 40vh;
-      }
-    }
 
-    .container {
+    .st-toolbar input {
-      flex: 1;
+      font: 12px var(--st-mono); color: var(--st-ink);
-      min-height: 0;
+      background: var(--st-surface); border: 1px solid var(--st-line); border-radius: 999px;
-      max-width: 1600px;
+      padding: 5px 12px; min-width: 160px;
-      width: 100%;
-      margin: 0 auto;
-      display: flex;
-      flex-direction: column;
-      gap: 15px;
-      overflow: hidden;
     }
+    .live-dot { display: inline-flex; align-items: center; gap: 6px; font-size: 12px; font-weight: 600; color: var(--st-ok); }
+    .live-dot::before { content: ""; width: 7px; height: 7px; border-radius: 50%; background: var(--st-ok); }
+    .live-dot.reconnecting { color: var(--st-crit); }
+    .live-dot.reconnecting::before { background: var(--st-crit); }
 
-    h1 {
+    /* ── host rows: 6 columns ── */
-      color: #333;
+    .hrow, .hhead {
-      margin-bottom: 5px;
+      display: grid;
-      margin-top: 15px; 
+      grid-template-columns: 175px 105px 70px 125px 85px minmax(0,1fr);
-      font-size: 1.5em;
+      gap: 4px 14px; align-items: center;
+      padding: 6px 14px; border-bottom: 1px solid var(--st-line-soft);
     }
-
+    .hrow:last-child { border-bottom: none; }
-    h2 {
+    .hrow:hover { background: var(--st-surface-2); }
-      color: #333;
+    .hhead {
-      margin-bottom: 10px;
+      font-family: var(--st-mono); font-size: 10.5px; text-transform: uppercase; letter-spacing: .05em;
-      font-size: 1.2em;
+      color: var(--st-faint); background: var(--st-surface-2); border-radius: 8px 8px 0 0;
-      padding: 10px 15px;
-      background: white;
-      border-radius: 6px;
-      box-shadow: 0 1px 4px rgba(0,0,0,0.1);
     }
+    .hhead .num, .hrow .num { text-align: right; }
 
-    .subtitle {
+    .hrow.sev-warn { border-left: 3px solid var(--st-warn); padding-left: 11px; }
-      color: #666;
+    .hrow.sev-crit { border-left: 3px solid var(--st-crit); padding-left: 11px; }
-      margin-bottom: 15px;
+
-      font-size: 0.9em;
+    .host { font-family: var(--st-mono); font-size: 13px; font-weight: 400; display: flex; align-items: center; gap: 7px; min-width: 0; }
+    .host.watched { font-weight: 700; }
+    .host a { color: inherit; text-decoration: none; overflow: hidden; text-overflow: ellipsis; white-space: nowrap; }
+    .host a:hover { color: var(--st-accent); }
+    .dot.down { background: var(--st-crit); box-shadow: 0 0 0 3px var(--st-crit-soft); }
+    .dot.overdue { background: transparent; border: 2px solid var(--st-warn); }
+    .stale { color: var(--st-faint); font-size: 10px; border: 1px solid var(--st-line); border-radius: 4px; padding: 1px 4px; font-family: var(--st-mono); white-space: nowrap; }
+    .flap { color: var(--st-warn); font-size: 10px; border: 1px solid var(--st-warn); border-radius: 4px; padding: 1px 4px; font-family: var(--st-mono); white-space: nowrap; }
+    .chip.outline { background: transparent; border: 1px solid var(--st-line); }
+
+    .lat { font-family: var(--st-mono); font-size: 12px; color: var(--st-muted); font-variant-numeric: tabular-nums; }
+    .since { font-family: var(--st-mono); font-size: 11.5px; color: var(--st-faint); font-variant-numeric: tabular-nums; }
+
+    .lastalert { font-family: var(--st-mono); font-size: 11.5px; color: var(--st-muted); overflow: hidden; text-overflow: ellipsis; white-space: nowrap; }
+    .lastalert .lvl { font-weight: 700; }
+    .lastalert .lvl.crit { color: var(--st-crit); }
+    .lastalert .lvl.warn { color: var(--st-warn); }
+    .lastalert .ago { color: var(--st-faint); }
+    .lastalert.quiet { color: var(--st-faint); }
+
+    /* ── hover card on the status chip ── */
+    .tip { position: relative; display: inline-flex; }
+    .tipbox {
+      display: none; position: absolute; z-index: 40; top: calc(100% + 6px); left: 0;
+      width: 340px; background: var(--st-surface); border: 1px solid var(--st-line); border-radius: 8px;
+      box-shadow: 0 8px 28px var(--shadow); padding: 10px 12px;
     }
+    .tip:hover .tipbox, .tip:focus-within .tipbox { display: block; }
+    .tipbox table { border-collapse: collapse; width: 100%; font-family: var(--st-mono); font-size: 11.5px; }
+    .tipbox th { text-align: left; font-size: 10px; text-transform: uppercase; letter-spacing: .05em; color: var(--st-faint); padding: 2px 6px 4px; border-bottom: 1px solid var(--st-line-soft); }
+    .tipbox td { padding: 3px 6px; color: var(--st-muted); }
+    .tipbox td:first-child { color: var(--st-faint); }
+    .tipbox .up { color: var(--st-ok); font-weight: 700; }
+    .tipbox .dn { color: var(--st-crit); font-weight: 700; }
+    .tipbox .ov { color: var(--st-warn); font-weight: 700; }
+    .tipbox .meta { margin-top: 6px; padding-top: 6px; border-top: 1px solid var(--st-line-soft); font-size: 11px; color: var(--st-faint); }
 
-    .content {
+    @media (max-width: 760px) {
-      display: flex;
+      .hhead { display: none; }
-      flex-direction: column;
+      .hrow { grid-template-columns: minmax(0,1fr) auto auto; padding: 8px 12px; }
-      gap: 15px;
+      .hrow .num { text-align: left; }
-    }
+      .hrow .since { display: none; }
-
+      .lastalert { grid-column: 1 / -1; }
-    .table-section {
+      .tipbox { left: auto; right: 0; }
-      background: white;
-      border-radius: 6px;
-      padding: 15px;
-      box-shadow: 0 1px 4px rgba(0,0,0,0.1);
-      overflow-x: auto;
-      overflow-y: auto;
-      max-height: 60vh;
-    }
-
-    .log-section {
-      flex: 1;
-      min-height: 0;
-      background: white;
-      border-radius: 6px;
-      padding: 15px;
-      box-shadow: 0 1px 4px rgba(0,0,0,0.1);
-      overflow-y: auto;
-    }
-
-    #ntable {
-      border-collapse: collapse;
-      width: 100%;
-      font-size: 0.9em;
-    }
-
-    #ntable td,
-    #ntable th {
-      border: 1px solid #e0e0e0;
-      text-align: left;
-      padding: 2px 4px;
-      white-space: nowrap;
-    }
-
-    #ntable tr:nth-child(even) {
-      background-color: #fafafa;
-    }
-
-    #ntable tr:hover {
-      background-color: #e3f2fd;
-    }
-
-    #ntable tbody tr.row-warning {
-      background-color: #fff8c5;
-    }
-
-    #ntable tbody tr.row-critical {
-      background-color: #fde8e8;
-    }
-
-    #ntable tbody tr.row-warning:hover {
-      background-color: #fff0a0;
-    }
-
-    #ntable tbody tr.row-critical:hover {
-      background-color: #f9c8c8;
-    }
-
-    #ntable th {
-      padding: 6px 8px;
-      background-color: #2196f3;
-      color: white;
-      font-weight: 600;
-      position: sticky;
-      top: 0;
-      z-index: 10;
-    }
-
-    #ntable
-      th:not(.sorttable_sorted):not(.sorttable_sorted_reverse):not(.sorttable_nosort):after {
-      content: " ⇅";
-      opacity: 0.5;
-    }
-    
-    /* Alert count column styling */
-    #ntable td.alert-warning {
-      color: #ff9800;
-      font-weight: bold;
-      text-align: center;
-    }
-    
-    #ntable td.alert-critical {
-      color: #f44336;
-      font-weight: bold;
-      text-align: center;
-    }
-
-    /* Scrollbar styling */
-    .log-section::-webkit-scrollbar {
-      width: 8px;
-    }
-
-    .log-section::-webkit-scrollbar-track {
-      background: #f1f1f1;
-      border-radius: 4px;
-    }
-
-    .log-section::-webkit-scrollbar-thumb {
-      background: #888;
-      border-radius: 4px;
-    }
-
-    .log-section::-webkit-scrollbar-thumb:hover {
-      background: #555;
-    }
-
-    /* Message styling */
-    #messages {
-      font-size: 0.85em;
-      line-height: 1.0;
-    }
-
-    #messages div {
-      padding: 5px 0;
-      border-bottom: 1px solid #f0f0f0;
-    }
-
-    /* Modal for connection status messages */
-    .connection-modal {
-      display: none;
-      position: fixed;
-      z-index: 1000;
-      left: 0;
-      top: 0;
-      width: 100%;
-      height: 100%;
-      background-color: rgba(0, 0, 0, 0.5);
-    }
-
-    .connection-modal.show {
-      display: flex;
-      justify-content: center;
-      align-items: center;
-    }
-
-    .connection-modal-content {
-      background-color: white;
-      padding: 30px 40px;
-      border-radius: 8px;
-      text-align: center;
-      box-shadow: 0 4px 20px rgba(0, 0, 0, 0.3);
-      min-width: 300px;
-    }
-
-    .connection-modal-content p {
-      margin: 0;
-      font-size: 16px;
-      color: #333;
-    }
-
-    /* State indicators */
-    .state-up {
-      color: #4caf50;
-      font-weight: 600;
-    }
-
-    .state-down {
-      color: #f44336;
-      font-weight: 700;
-    }
-
-    .state-overdue {
-      color: #ff9800;
-      font-weight: 700;
     }
   </style>
-  <script type="text/javascript">
-    var cnt = 0;
-    var nTable = document;
-    var name_idx = {};
-    var c = 0;
-    var HBD_VERSION = "{{ hbd_version }}";
 
-    function hostNameHtml(data) {
-      var nameHtml = data.name;
-      if (!data.hbc_version || data.hbc_version !== HBD_VERSION) {
-        nameHtml += ' 🥀';
-      }
-      return data.dyn ? '<b>' + nameHtml + '</b>' : nameHtml;
-    }
-
-    function setup() {
-      name_idx = {};
-      nTable = document.getElementById("ntable");
-      for (var i = 0, row; (row = nTable.rows[i]); i++) {
-        if (i == 0) continue;
-        var cell = nTable.rows[i].cells[0];
-        var name = cell.dataset.name || cell.innerText.replace(/\s*🥀\s*$/, '').trim();
-        name_idx[name] = nTable.rows[i];
-      }
-    }
-
-    function updateRowAlert(row, data) {
-      var criticalUnacked = data.alert_critical_unacked || 0;
-      var criticalAcked = data.alert_critical_acked || 0;
-      var warningUnacked = data.alert_warning_unacked || 0;
-      var warningAcked = data.alert_warning_acked || 0;
-      row.classList.remove('row-warning', 'row-critical');
-      if (criticalUnacked > 0 || criticalAcked > 0) {
-        row.classList.add('row-critical');
-      } else if (warningUnacked > 0 || warningAcked > 0) {
-        row.classList.add('row-warning');
-      }
-    }
-
-    function createRow(data) {
-      var row = document.createElement("tr");
-      var c_name = document.createElement("td");
-      var c_warning = document.createElement("td");
-      c_warning.style.textAlign = "center";
-      c_warning.style.color = "#ff9800";
-      c_warning.style.fontWeight = "bold";
-      var c_critical = document.createElement("td");
-      c_critical.style.textAlign = "center";
-      c_critical.style.color = "#f44336";
-      c_critical.style.fontWeight = "bold";
-      var c_ipv4addr = document.createElement("td");
-      var c_ipv4state = document.createElement("td");
-      var c_ipv4latency = document.createElement("td");
-      c_ipv4latency.style.textAlign = "right";
-      var c_ipv4statets = document.createElement("td");
-      c_ipv4statets.style.textAlign = "right";
-      var c_ipv6addr = document.createElement("td");
-      var c_ipv6state = document.createElement("td");
-      var c_ipv6latency = document.createElement("td");
-      c_ipv6latency.style.textAlign = "right";
-      var c_ipv6statets = document.createElement("td");
-      c_ipv6statets.style.textAlign = "right";
-      row.appendChild(c_name);
-      row.appendChild(c_warning);
-      row.appendChild(c_critical);
-      row.appendChild(c_ipv4addr);
-      row.appendChild(c_ipv4state);
-      row.appendChild(c_ipv4latency);
-      row.appendChild(c_ipv4statets);
-      row.appendChild(c_ipv6addr);
-      row.appendChild(c_ipv6state);
-      row.appendChild(c_ipv6latency);
-      row.appendChild(c_ipv6statets);
-      c_name.dataset.name = data.name;
-      c_name.innerHTML = hostNameHtml(data);
-      
-      // Set alert counts in "x/y" format (unacked/acked)
-      var warningUnacked = data.alert_warning_unacked || 0;
-      var warningAcked = data.alert_warning_acked || 0;
-      var criticalUnacked = data.alert_critical_unacked || 0;
-      var criticalAcked = data.alert_critical_acked || 0;
-      
-      if (warningUnacked > 0 || warningAcked > 0) {
-        c_warning.innerHTML = warningAcked > 0 ? warningUnacked + "/" + warningAcked : warningUnacked;
-      } else {
-        c_warning.innerHTML = "";
-      }
-      
-      if (criticalUnacked > 0 || criticalAcked > 0) {
-        c_critical.innerHTML = criticalAcked > 0 ? criticalUnacked + "/" + criticalAcked : criticalUnacked;
-      } else {
-        c_critical.innerHTML = "";
-      }
-      
-      c_ipv4addr.innerHTML = data.connections[0].addr;
-      c_ipv4state.innerHTML = data.connections[0].state;
-      if (data.connections.length > 1) {
-        c_ipv6addr.innerHTML = data.connections[1].addr;
-        c_ipv6state.innerHTML = data.connections[1].state;
-      }
-      var table = document.getElementById("ntablebody"); // find table to append to
-      table.appendChild(row); // append row to table
-      name_idx[c_name] = row;
-      updateRowAlert(row, data);
-    }
-
-    function formatTS(ts) {
-      const now = new Date();
-      const d = new Date(ts * 1000);
-
-      const pad = n => String(n).padStart(2, '0');
-      const timeStr = `${pad(d.getHours())}:${pad(d.getMinutes())}:${pad(d.getSeconds())}`;
-
-      // Same calendar day → show time only
-      if (d.toDateString() === now.toDateString()) {
-        return timeStr;
-      }
-
-      // Within 8 days → show "-X d hh:mm:ss"
-      const todayStart = new Date(now.getFullYear(), now.getMonth(), now.getDate());
-      const dStart = new Date(d.getFullYear(), d.getMonth(), d.getDate());
-      const diffDays = Math.round((todayStart - dStart) / 86400000);
-      if (diffDays < 8) {
-        return `-${diffDays}d ${timeStr}`;
-      }
-
-      // Older → date only
-      return `${d.getFullYear()}-${pad(d.getMonth() + 1)}-${pad(d.getDate())}`;
-    }
-
-    function update_table(data) {
-      if (!(data.name in name_idx)) {
-        createRow(data);
-        setup();
-      }
-      
-      // Update name cell (version indicator)
-      var nameCell = name_idx[data.name].cells[0];
-      nameCell.dataset.name = data.name;
-      nameCell.innerHTML = hostNameHtml(data);
-
-      // Update warning and critical counts in "x/y" format (unacked/acked)
-      var warningUnacked = data.alert_warning_unacked || 0;
-      var warningAcked = data.alert_warning_acked || 0;
-      var criticalUnacked = data.alert_critical_unacked || 0;
-      var criticalAcked = data.alert_critical_acked || 0;
-      
-      if (warningUnacked > 0 || warningAcked > 0) {
-        name_idx[data.name].cells[1].innerHTML = warningAcked > 0 ? warningUnacked + "/" + warningAcked : warningUnacked;
-      } else {
-        name_idx[data.name].cells[1].innerHTML = "";
-      }
-      
-      if (criticalUnacked > 0 || criticalAcked > 0) {
-        name_idx[data.name].cells[2].innerHTML = criticalAcked > 0 ? criticalUnacked + "/" + criticalAcked : criticalUnacked;
-      } else {
-        name_idx[data.name].cells[2].innerHTML = "";
-      }
-
-      for  (var i = 0; i < data.connections.length; i++) {
-        // Offset by 2 for the warning/critical count columns
-        name_idx[data.name].cells[3 + i * 4].innerHTML = data.connections[i].addr;
-        name_idx[data.name].cells[6 + i * 4].innerHTML = formatTS(
-          data.connections[i].statetime
-        );
-        if (data.connections[i].state == "up") {
-          state = '<span class="state-up">up</span>';
-          latency = Number.parseFloat(data.connections[i].rtts[0]).toFixed(2);
-        } else {
-          if (data.connections[i].state == "unknown") {
-            state = "";
-            latency = "";
-            name_idx[data.name].cells[3 + i * 4].innerHTML = "";
-            name_idx[data.name].cells[6 + i * 4].innerHTML = "";
-          } else if (data.connections[i].state == "down") {
-            state = '<span class="state-down">down</span>';
-            latency = "-";
-          } else if (data.connections[i].state == "overdue") {
-            state = '<span class="state-overdue">overdue</span>';
-            latency = "-";
-          } else {
-            state = "<b>" + data.connections[i].state + "</b>";
-            latency = "-";
-          }
-        }
-        name_idx[data.name].cells[4 + i * 4].innerHTML = state;
-        name_idx[data.name].cells[5 + i * 4].innerHTML = latency;
-      }
-      updateRowAlert(name_idx[data.name], data);
-    }
-
-    function WS_Connect() {
-      if ("WebSocket" in window) {
-        //N.B: subprotocol field causes chrome to error 1006
-        var ws_hbd = new WebSocket("{{heartbeat_ws_url}}", /* "hdb" */ );
-
-        ws_hbd.onopen = function () {
-          // Web Socket is connected, send data using send()
-          console.log("ws connect {{heartbeat_ws_url}}");
-          // Hide modal window if visible
-          var modal = document.getElementById("connectionModal");
-          if (modal) {
-            modal.classList.remove("show");
-          }
-          ws_hbd.send("heartbeat_web");
-        };
-
-        ws_hbd.onerror = function (event) {
-          console.log(event);
-        };
-
-        ws_hbd.onmessage = function (event) {
-          /*      console.log(event.data); */
-          var state = JSON.parse(event.data);
-          /* console.log("State: " + state.type); */
-          if (state.type == "host") {
-            update_table(state.data);
-          } else if (state.type == "message") {
-            var msgs = document.getElementById("messages");
-            msgs.insertAdjacentHTML("afterbegin", "<div>" + state.data + "</div>");
-          }
-          cnt++;
-        };
-
-        ws_hbd.onclose = function (event) {
-          /*     console.log(event); */
-          console.log("Connection is closed, reopening");
-          // Show modal window
-          var modal = document.getElementById("connectionModal");
-          if (modal) {
-            modal.classList.add("show");
-          }
-          setTimeout(function () {
-            WS_Connect();
-          }, 3000);
-        };
-      } else {
-        // The browser doesn't support WebSocket
-        console.log("WebSocket NOT supported by your Browser!");
-      }
-    }
-    WS_Connect();
-  </script>
   <body>
     {% include 'nav.html' %}
 
-    {% include 'menu.html' %}
+    <div class="st-toolbar">
-
+      <span class="brand">hbd<span class="tld">·live</span></span>
-    <div class="container">
+      <span class="hintline">every host, one line — details on hover, history in the overview</span>
-      <div>
+      <span class="spacer"></span>
-        <h1>{{ header }}</h1>
+      <input type="text" id="host-filter" placeholder="host filter (regex)">
-        <p class="subtitle">Real-time host monitoring and event log</p>
+      <span class="live-dot reconnecting" id="ws-status">connecting…</span>
-      </div>
-      
-      <div class="table-section">
-        <table id="ntable" class="sortable">
-          <thead>
-            <tr>
-              <th>Name</th>
-              <th style="text-align: center" title="Warning Alerts">⚠️</th>
-              <th style="text-align: center" title="Critical Alerts">🔴</th>
-              <th>IPv4 Addr</th>
-              <th>State</th>
-              <th style="text-align: right">Latency</th>
-              <th style="text-align: right">Last State</th>
-              <th>IPv6 Addr</th>
-              <th>State</th>
-              <th style="text-align: right">Latency</th>
-              <th style="text-align: right">Last State</th>
-            </tr>
-          </thead>
-          <tbody id="ntablebody">
-            {% for host in hosts %}
-            <tr class="{% if host.alert_critical_unacked > 0 or host.alert_critical_acked > 0 %}row-critical{% elif host.alert_warning_unacked > 0 or host.alert_warning_acked > 0 %}row-warning{% endif %}">
-              <td data-name="{{ host.name }}">{{ host.name }}{% if not host.hbc_version or host.hbc_version != hbd_version %} 🥀{% endif %}</td>
-              <td style="text-align: center; color: #ff9800; font-weight: bold;">
-                {%- set warning_unacked = host.alert_warning_unacked -%}
-                {%- set warning_acked = host.alert_warning_acked -%}
-                {%- if warning_unacked > 0 or warning_acked > 0 -%}
-                  {{ warning_unacked }}{% if warning_acked > 0 %}/{{ warning_acked }}{% endif %}
-                {%- endif -%}
-              </td>
-              <td style="text-align: center; color: #f44336; font-weight: bold;">
-                {%- set critical_unacked = host.alert_critical_unacked -%}
-                {%- set critical_acked = host.alert_critical_acked -%}
-                {%- if critical_unacked > 0 or critical_acked > 0 -%}
-                  {{ critical_unacked }}{% if critical_acked > 0 %}/{{ critical_acked }}{% endif %}
-                {%- endif -%}
-              </td>
-              {% for conn in host.connections %}
-                <td>{{ conn.addr if conn.addr else '' }}</td>
-                <td>{{ conn.state if conn.state else '' }}</td>
-                <td style="text-align: right">{{ conn.latency if conn.latency else '' }}</td>
-                <td style="text-align: right">{{ conn.last_state_ts if conn.last_state_ts else '' }}</td>
-              {% endfor %}
-              {% if host.connections|length == 0 %}
-                <td></td><td></td><td></td><td></td>
-                <td></td><td></td><td></td><td></td>
-              {% elif host.connections|length == 1 %}
-                <td></td><td></td><td></td><td></td>
-              {% endif %}
-            </tr>
-            {% endfor %}
-          </tbody>
-        </table>
-      </div>
-      
-      <div class="log-section">
-        <h2>Log of Events</h2>
-        <div id="messages"></div>
-      </div>
     </div>
+    <svg class="pulse" viewBox="0 0 1200 14" preserveAspectRatio="none" aria-hidden="true">
+      <polyline points="0,10 340,10 352,10 358,3 364,13 370,1 378,12 384,10 560,10 572,10 578,3 584,13 590,1 598,12 604,10 1200,10"/>
+    </svg>
 
-    {% include 'foot.html' %}
+    <div class="live-main">
-    
+      <section class="st">
-    <!-- Connection status modal -->
+        <div class="sec-head">
-    <div id="connectionModal" class="connection-modal">
+          <h2>hosts<span class="colon">:</span></h2>
-      <div class="connection-modal-content">
+          <span class="count" id="host-counts"></span>
-        <p>⚠️ Connection is closed, reopening...</p>
+          <span class="sub">sorted by severity, then name</span>
-      </div>
+        </div>
+        <div class="list">
+          <div class="hhead">
+            <span>host</span><span>status</span><span class="num">latency</span>
+            <span>alerts</span><span>since</span><span>last alert</span>
+          </div>
+          <div id="host-rows"></div>
+        </div>
+      </section>
     </div>
 
     <script>
-      setup();
+      var HBD_VERSION = "{{ hbd_version }}";
+      var INITIAL_HOSTS = {{ hosts | tojson }};
+
+      var hostData = {};    // name -> latest stateinfo payload
+      var lastAlert = {};   // name -> {level, message, ts}
+      var hostFilterRe = null;
+
+      function escHtml(s) {
+        return String(s).replace(/&/g,'&amp;').replace(/</g,'&lt;').replace(/>/g,'&gt;').replace(/"/g,'&quot;').replace(/'/g,'&#39;');
+      }
+
+      // ---- derived per-host facts ----------------------------------------
+      var STATE_RANK = { down: 3, overdue: 2, up: 1 };
+
+      function worstConnection(data) {
+        var worst = null;
+        (data.connections || []).forEach(function (c) {
+          if (!c.state || c.state === 'unknown') return;
+          if (!worst || (STATE_RANK[c.state] || 0) > (STATE_RANK[worst.state] || 0)) worst = c;
+        });
+        return worst;
+      }
+
+      function severity(data) {
+        var w = worstConnection(data);
+        if ((data.alert_critical_unacked || 0) > 0 || (data.alert_critical_acked || 0) > 0) return 'crit';
+        if (w && w.state === 'down') return 'crit';
+        if ((data.alert_warning_unacked || 0) > 0 || (data.alert_warning_acked || 0) > 0) return 'warn';
+        if (w && w.state === 'overdue') return 'warn';
+        return 'ok';
+      }
+      var SEV_ORDER = { crit: 0, warn: 1, ok: 2 };
+
+      function relTime(ts) {
+        if (!ts) return '—';
+        var s = Math.max(0, Math.floor(Date.now() / 1000 - ts));
+        if (s < 60) return s + 's';
+        if (s < 3600) return Math.floor(s / 60) + 'm';
+        if (s < 86400) return Math.floor(s / 3600) + 'h ' + Math.floor((s % 3600) / 60) + 'm';
+        if (s < 86400 * 30) return Math.floor(s / 86400) + 'd ' + Math.floor((s % 86400) / 3600) + 'h';
+        return Math.floor(s / 86400) + 'd';
+      }
+
+      function fmtClock(ts) {
+        var d = new Date(ts * 1000);
+        function p(n) { return n < 10 ? '0' + n : '' + n; }
+        var t = p(d.getHours()) + ':' + p(d.getMinutes()) + ':' + p(d.getSeconds());
+        return (d.toDateString() === new Date().toDateString())
+          ? t
+          : d.getFullYear() + '-' + p(d.getMonth() + 1) + '-' + p(d.getDate()) + ' ' + t;
+      }
+
+      // ---- host row rendering ---------------------------------------------
+      function statusChip(data) {
+        var w = worstConnection(data);
+        if (!w) return '<span class="chip outline">—</span>';
+        var cls = w.state === 'up' ? 'ok' : (w.state === 'down' ? 'crit' : 'warn');
+        return '<span class="chip ' + cls + '">' + escHtml(w.state) + '</span>';
+      }
+
+      function tipCard(data) {
+        var rows = '';
+        (data.connections || []).forEach(function (c) {
+          if (!c.state || c.state === 'unknown') return;
+          var cls = c.state === 'up' ? 'up' : (c.state === 'down' ? 'dn' : 'ov');
+          var fam = (c.afam || '').replace('IPv', 'v');
+          var rtt = (c.state === 'up' && c.rtts && c.rtts[0]) ? Math.round(c.rtts[0]) + ' ms' : '—';
+          rows += '<tr><td>' + escHtml(fam) + '</td><td>' + escHtml(c.addr || '—') + '</td>'
+                + '<td class="' + cls + '">' + escHtml(c.state) + '</td>'
+                + '<td>' + rtt + '</td><td>' + (c.statetime ? fmtClock(c.statetime) : '—') + '</td></tr>';
+        });
+        if (!rows) rows = '<tr><td colspan="5">no connections yet</td></tr>';
+        var meta = [];
+        meta.push('agent ' + (data.hbc_version ? 'v' + escHtml(String(data.hbc_version)) : 'unknown'));
+        if (data.dyn) meta.push('dyndns');
+        if (data.watched) meta.push('watched');
+        if (data.interval) meta.push('interval ' + escHtml(String(data.interval)) + ' s');
+        return '<span class="tipbox" role="tooltip">'
+          + '<table><thead><tr><th></th><th>address</th><th>state</th><th>rtt</th><th>since</th></tr></thead>'
+          + '<tbody>' + rows + '</tbody></table>'
+          + '<div class="meta">' + meta.join(' · ') + '</div></span>';
+      }
+
+      function alertChips(data) {
+        var cu = data.alert_critical_unacked || 0, ca = data.alert_critical_acked || 0;
+        var wu = data.alert_warning_unacked || 0, wa = data.alert_warning_acked || 0;
+        var out = '';
+        if (cu) out += '<span class="chip crit"><b>' + cu + '</b> crit</span> ';
+        if (wu) out += '<span class="chip warn"><b>' + wu + '</b> warn</span> ';
+        var acked = ca + wa;
+        if (acked) out += '<span class="chip outline">' + acked + ' ack’d</span>';
+        return out;
+      }
+
+      function lastAlertHtml(name) {
+        var a = lastAlert[name];
+        if (!a) return '<span class="lastalert quiet">—</span>';
+        var lvl = (a.level || '').toUpperCase();
+        var quiet = (lvl !== 'WARNING' && lvl !== 'CRITICAL');
+        var lvlCls = lvl === 'CRITICAL' ? 'crit' : (lvl === 'WARNING' ? 'warn' : '');
+        var short = lvl === 'CRITICAL' ? 'CRIT' : (lvl === 'WARNING' ? 'WARN' : escHtml(lvl));
+        return '<span class="lastalert' + (quiet ? ' quiet' : '') + '" title="' + escHtml(a.message) + '">'
+          + '<span class="lvl ' + lvlCls + '">' + short + '</span> '
+          + escHtml(a.message)
+          + (a.ts ? ' <span class="ago">· ' + relTime(a.ts) + ' ago</span>' : '')
+          + '</span>';
+      }
+
+      function renderRow(data) {
+        var name = data.name;
+        var sev = severity(data);
+        var row = document.querySelector('.hrow[data-host="' + CSS.escape(name) + '"]');
+        var isNew = false;
+        if (!row) {
+          row = document.createElement('div');
+          row.dataset.host = name;
+          isNew = true;
+        }
+        row.className = 'hrow' + (sev !== 'ok' ? ' sev-' + sev : '');
+        row.dataset.sev = sev;
+
+        var w = worstConnection(data);
+        var up = (data.connections || []).find(function (c) { return c.state === 'up' && c.rtts && c.rtts.length; });
+        var lat = up ? Math.round(up.rtts[0]) + ' ms' : '—';
+        var since = w ? relTime(w.statetime) : '—';
+        var staleTag = '';
+        if (!data.hbc_version) staleTag = '<span class="stale" title="agent version unknown">agent?</span>';
+        else if (data.hbc_version !== HBD_VERSION) {
+          staleTag = '<span class="stale" title="agent v' + escHtml(String(data.hbc_version)) + ' ≠ server v' + HBD_VERSION + '">v' + escHtml(String(data.hbc_version)) + '</span>';
+        }
+        var dotCls = w ? (w.state === 'up' ? sev : (w.state === 'down' ? 'down' : 'overdue')) : 'off';
+        var flapTag = '';
+        var flapping = data.flapping || [];
+        if (flapping.length) {
+          var svcs = flapping.map(function (s) { return s || 'host'; }).join(', ');
+          flapTag = '<span class="flap" title="notifications suppressed: ' + escHtml(svcs) + '">flapping</span>';
+        }
+
+        row.innerHTML =
+          '<span class="host' + (data.watched ? ' watched' : '') + '">'
+          + '<span class="dot ' + dotCls + '"></span>'
+          + '<a href="/plugins#' + encodeURIComponent(name) + '">' + escHtml(name) + '</a>' + staleTag + flapTag + '</span>'
+          + '<span class="tip" tabindex="0">' + statusChip(data) + tipCard(data) + '</span>'
+          + '<span class="lat num">' + lat + '</span>'
+          + '<span>' + alertChips(data) + '</span>'
+          + '<span class="since">' + since + '</span>'
+          + lastAlertHtml(name);
+
+        if (isNew) {
+          document.getElementById('host-rows').appendChild(row);
+          sortRows();
+        }
+        applyHostFilter(row);
+      }
+
+      function sortRows() {
+        var container = document.getElementById('host-rows');
+        var rows = [...container.children];
+        rows.sort(function (a, b) {
+          var d = (SEV_ORDER[a.dataset.sev] ?? 2) - (SEV_ORDER[b.dataset.sev] ?? 2);
+          return d !== 0 ? d : a.dataset.host.localeCompare(b.dataset.host);
+        });
+        rows.forEach(function (r) { container.appendChild(r); });
+      }
+
+      function updateCounts() {
+        var total = 0, up = 0, overdue = 0, down = 0;
+        Object.values(hostData).forEach(function (d) {
+          total++;
+          var w = worstConnection(d);
+          if (!w) return;
+          if (w.state === 'up') up++;
+          else if (w.state === 'overdue') overdue++;
+          else if (w.state === 'down') down++;
+        });
+        var parts = [total + ' hosts', up + ' up'];
+        if (overdue) parts.push(overdue + ' overdue');
+        if (down) parts.push(down + ' down');
+        document.getElementById('host-counts').textContent = parts.join(' · ');
+      }
+
+      function updateHost(data) {
+        var prev = hostData[data.name];
+        var prevSev = prev ? severity(prev) : null;
+        hostData[data.name] = data;
+        renderRow(data);
+        if (prevSev !== severity(data)) sortRows();
+        updateCounts();
+      }
+
+      // ---- host filter -----------------------------------------------------
+      function applyHostFilter(row) {
+        var rows = row ? [row] : [...document.querySelectorAll('.hrow')];
+        rows.forEach(function (r) {
+          r.style.display = (!hostFilterRe || hostFilterRe.test(r.dataset.host)) ? '' : 'none';
+        });
+      }
+      document.getElementById('host-filter').addEventListener('input', function () {
+        var v = this.value.trim();
+        try {
+          hostFilterRe = v ? new RegExp(v, 'i') : null;
+          this.style.borderColor = '';
+        } catch (_) {
+          hostFilterRe = null;
+          this.style.borderColor = 'var(--st-crit)';
+        }
+        applyHostFilter();
+      });
+
+      // ---- last alert tracking ---------------------------------------------
+      var ALERT_LEVELS = { WARNING: 1, CRITICAL: 1, RECOVER: 1 };
+
+      function noteAlert(host, level, message, ts) {
+        if (!host || !ALERT_LEVELS[(level || '').toUpperCase()]) return;
+        var cur = lastAlert[host];
+        if (cur && cur.ts && ts && ts < cur.ts) return;  // keep the newest
+        lastAlert[host] = { level: level, message: message, ts: ts };
+        if (hostData[host]) renderRow(hostData[host]);
+      }
+
+      async function seedActiveAlerts() {
+        try {
+          var r = await fetch('/api/0/alerts');
+          if (!r.ok) return;
+          var data = await r.json();
+          (data.alerts || []).forEach(function (a) {
+            var metric = (a.metric_path.includes('.')
+              ? a.metric_path.slice(a.metric_path.indexOf('.') + 1)
+              : a.metric_path).replace(/_status_code$/, '');
+            var msg = metric + ' = ' + a.last_value
+              + (a.formatted_message ? ' ' + a.formatted_message
+                 : (a.operator && a.threshold_value != null ? ' (' + a.operator + ' ' + a.threshold_value + ')' : ''));
+            noteAlert(a.hostname, a.level, msg, a.since);
+          });
+        } catch (_) { /* dashboard still works without the seed */ }
+      }
+
+      // ---- websocket -----------------------------------------------------------
+      function setWsStatus(ok) {
+        var el = document.getElementById('ws-status');
+        el.classList.toggle('reconnecting', !ok);
+        el.textContent = ok ? 'live' : 'reconnecting…';
+      }
+
+      function WS_Connect() {
+        if (!("WebSocket" in window)) {
+          console.log("WebSocket NOT supported by your Browser!");
+          return;
+        }
+        var ws_hbd = new WebSocket("{{ heartbeat_ws_url }}");
+
+        ws_hbd.onopen = function () {
+          setWsStatus(true);
+          ws_hbd.send("heartbeat_web");
+        };
+
+        ws_hbd.onerror = function (event) { console.log(event); };
+
+        ws_hbd.onmessage = function (event) {
+          var state = JSON.parse(event.data);
+          if (state.type === "host") {
+            updateHost(state.data);
+          } else if (state.type === "message") {
+            var msg = state.data;
+            noteAlert(msg.host, msg.level, msg.message, msg.ts);
+          }
+        };
+
+        ws_hbd.onclose = function () {
+          setWsStatus(false);
+          setTimeout(WS_Connect, 3000);
+        };
+      }
+
+      // ---- boot -----------------------------------------------------------------
+      INITIAL_HOSTS.forEach(function (h) { hostData[h.name] = h; });
+      Object.values(hostData).forEach(renderRow);
+      sortRows();
+      updateCounts();
+      seedActiveAlerts();
+      WS_Connect();
+
+      // refresh relative times once a minute
+      setInterval(function () {
+        Object.values(hostData).forEach(renderRow);
+      }, 60000);
     </script>
   </body>
 </html>

hbd/server/templates/log.html

@@ -0,0 +1,228 @@
+<!DOCTYPE html>
+<html>
+  {% include 'head.html' %}
+  <link rel="stylesheet" href="/static/hbd-ui.css">
+  <script src="/static/hbd-ui.js"></script>
+
+  <style>
+    html, body { height: auto; overflow: visible; }
+    body { background: var(--st-paper); padding: 60px 0 0; }
+
+    .log-main { max-width: 1100px; margin: 0 auto; padding: 0 20px 60px; }
+
+    .st-toolbar input, .st-toolbar select {
+      font: 12px var(--st-mono); color: var(--st-ink);
+      background: var(--st-surface); border: 1px solid var(--st-line); border-radius: 999px;
+      padding: 5px 12px;
+    }
+    .live-dot { display: inline-flex; align-items: center; gap: 6px; font-size: 12px; font-weight: 600; color: var(--st-ok); }
+    .live-dot::before { content: ""; width: 7px; height: 7px; border-radius: 50%; background: var(--st-ok); }
+    .live-dot.reconnecting { color: var(--st-crit); }
+    .live-dot.reconnecting::before { background: var(--st-crit); }
+
+    .logrow {
+      display: grid; grid-template-columns: 130px 74px 150px minmax(0,1fr);
+      gap: 4px 14px; align-items: baseline;
+      padding: 4px 14px; border-bottom: 1px solid var(--st-line-soft);
+      font-family: var(--st-mono); font-size: 11.5px;
+    }
+    .logrow:last-child { border-bottom: none; }
+    .logrow .t { color: var(--st-faint); font-variant-numeric: tabular-nums; white-space: nowrap; }
+    .logrow .h { color: var(--st-ink); font-weight: 600; overflow: hidden; text-overflow: ellipsis; white-space: nowrap; }
+    .logrow .m { color: var(--st-muted); overflow: hidden; text-overflow: ellipsis; white-space: nowrap; }
+
+    .empty { padding: 24px 14px; text-align: center; color: var(--st-faint); font-size: 12.5px; }
+    .loadmore {
+      display: block; margin: 14px auto 0; cursor: pointer;
+      font-family: var(--st-mono); font-size: 12px; color: var(--st-muted);
+      background: var(--st-surface); border: 1px solid var(--st-line); border-radius: 999px;
+      padding: 6px 16px;
+    }
+    .loadmore:hover { color: var(--st-accent); border-color: var(--st-accent); }
+    .loadmore:disabled { opacity: .5; cursor: default; }
+
+    @media (max-width: 760px) {
+      .logrow { grid-template-columns: 66px minmax(0,1fr); }
+      .logrow .lvlc, .logrow .h { display: none; }
+    }
+  </style>
+
+  <body>
+    {% include 'nav.html' %}
+
+    <div class="st-toolbar">
+      <span class="brand">hbd<span class="tld">·log</span></span>
+      <span class="hintline">connectivity and alert events — full journal history</span>
+      <span class="spacer"></span>
+      <input type="text" id="filter-host" placeholder="host">
+      <select id="filter-level">
+        <option value="">level</option>
+        <option value="info">INFO</option>
+        <option value="warning">WARNING</option>
+        <option value="critical">CRITICAL</option>
+        <option value="recover">RECOVER</option>
+        <option value="unknown">UNKNOWN</option>
+      </select>
+      <input type="text" id="filter-msg" placeholder="message">
+      <span class="live-dot reconnecting" id="ws-status">connecting…</span>
+    </div>
+    <svg class="pulse" viewBox="0 0 1200 14" preserveAspectRatio="none" aria-hidden="true">
+      <polyline points="0,10 340,10 352,10 358,3 364,13 370,1 378,12 384,10 560,10 572,10 578,3 584,13 590,1 598,12 604,10 1200,10"/>
+    </svg>
+
+    <div class="log-main">
+      <section class="st">
+        <div class="sec-head">
+          <h2>log<span class="colon">:</span></h2>
+          <span class="count" id="log-count"></span>
+        </div>
+        <div class="list" id="messages">
+          <div class="empty">Loading events…</div>
+        </div>
+        <button class="loadmore" id="load-older" style="display:none">load older</button>
+      </section>
+    </div>
+
+    <script>
+      var PAGE_SIZE = 200;
+      var oldestTs = null;   // ts of the oldest loaded row (pagination cursor)
+      var seq = 0;           // guards against out-of-order fetch responses
+      var newestTs = null;   // ts of the newest loaded row (dedup guard for the WS tail)
+
+      function escHtml(s) {
+        return String(s).replace(/&/g,'&amp;').replace(/</g,'&lt;').replace(/>/g,'&gt;').replace(/"/g,'&quot;').replace(/'/g,'&#39;');
+      }
+
+      function fmtClock(ts) {
+        var d = new Date(ts * 1000);
+        function p(n) { return n < 10 ? '0' + n : '' + n; }
+        var t = p(d.getHours()) + ':' + p(d.getMinutes()) + ':' + p(d.getSeconds());
+        return (d.toDateString() === new Date().toDateString())
+          ? t
+          : d.getFullYear() + '-' + p(d.getMonth() + 1) + '-' + p(d.getDate()) + ' ' + t;
+      }
+
+      function logRowHtml(msg) {
+        var lvl = (msg.level || 'INFO');
+        var lc = lvl.toLowerCase();
+        var chipCls = lc === 'critical' ? 'crit' : (lc === 'warning' ? 'warn' : (lc === 'recover' ? 'ok' : ''));
+        var short = lc === 'critical' ? 'CRIT' : (lc === 'warning' ? 'WARN' : escHtml(lvl));
+        return '<div class="logrow">'
+          + '<span class="t">' + fmtClock(msg.ts) + '</span>'
+          + '<span class="lvlc"><span class="chip ' + chipCls + '">' + short + '</span></span>'
+          + '<span class="h">' + escHtml(msg.host || '') + (msg.service ? ' · ' + escHtml(msg.service) : '') + '</span>'
+          + '<span class="m">' + escHtml(msg.message || '') + '</span>'
+          + '</div>';
+      }
+
+      function currentFilters() {
+        return {
+          host: document.getElementById('filter-host').value.trim(),
+          level: document.getElementById('filter-level').value,
+          q: document.getElementById('filter-msg').value.trim()
+        };
+      }
+
+      function matchesFilters(msg) {
+        var f = currentFilters();
+        if (f.host && !(msg.host || '').toLowerCase().includes(f.host.toLowerCase())) return false;
+        if (f.level && (msg.level || '').toLowerCase() !== f.level) return false;
+        if (f.q && !(msg.message || '').toLowerCase().includes(f.q.toLowerCase())) return false;
+        return true;
+      }
+
+      function updateCount() {
+        var n = document.querySelectorAll('#messages .logrow').length;
+        document.getElementById('log-count').textContent = n ? n + ' events loaded' : '';
+      }
+
+      async function fetchLog(reset) {
+        var mySeq = ++seq;
+        var box = document.getElementById('messages');
+        var btn = document.getElementById('load-older');
+        btn.disabled = true;
+        var f = currentFilters();
+        var params = new URLSearchParams();
+        params.set('limit', String(PAGE_SIZE));
+        if (f.host) params.set('host', f.host);
+        if (f.level) params.set('level', f.level);
+        if (f.q) params.set('q', f.q);
+        if (!reset && oldestTs != null) params.set('before', String(oldestTs));
+        var data;
+        try {
+          var r = await fetch('/api/0/log?' + params.toString());
+          if (!r.ok) throw new Error('HTTP ' + r.status);
+          data = await r.json();
+        } catch (e) {
+          if (mySeq === seq && reset) {
+            box.innerHTML = '<div class="empty">Failed to load events: ' + escHtml(e.message) + '</div>';
+          }
+          btn.disabled = false;
+          return;
+        }
+        if (mySeq !== seq) return;  // a newer request superseded this one
+        if (reset) { box.innerHTML = ''; oldestTs = null; }
+        var html = '';
+        data.events.forEach(function (ev) {
+          html += logRowHtml(ev);
+          if (ev.ts != null && (oldestTs == null || ev.ts < oldestTs)) oldestTs = ev.ts;
+          if (ev.ts != null && (newestTs == null || ev.ts > newestTs)) newestTs = ev.ts;
+        });
+        box.insertAdjacentHTML('beforeend', html);
+        if (!box.children.length) box.innerHTML = '<div class="empty">No events found.</div>';
+        btn.style.display = data.more ? '' : 'none';
+        btn.disabled = false;
+        updateCount();
+      }
+
+      function onFilterChange() { fetchLog(true); }
+
+      function debounce(fn, ms) {
+        var t;
+        return function () { clearTimeout(t); t = setTimeout(fn, ms); };
+      }
+      document.getElementById('filter-host').addEventListener('input', debounce(onFilterChange, 300));
+      document.getElementById('filter-msg').addEventListener('input', debounce(onFilterChange, 300));
+      document.getElementById('filter-level').addEventListener('change', onFilterChange);
+      document.getElementById('load-older').addEventListener('click', function () { fetchLog(false); });
+
+      // ---- live tail over the shared websocket --------------------------------
+      function setWsStatus(ok) {
+        var el = document.getElementById('ws-status');
+        el.classList.toggle('reconnecting', !ok);
+        el.textContent = ok ? 'live' : 'reconnecting…';
+      }
+
+      function WS_Connect() {
+        if (!("WebSocket" in window)) return;
+        var ws_hbd = new WebSocket("{{ heartbeat_ws_url }}");
+        ws_hbd.onopen = function () {
+          setWsStatus(true);
+          ws_hbd.send("heartbeat_web");
+        };
+        ws_hbd.onmessage = function (event) {
+          var state = JSON.parse(event.data);
+          // history replays are covered by the API seed; only tail new events
+          if (state.type !== "message" || state.history) return;
+          var msg = state.data;
+          if (!matchesFilters(msg)) return;
+          if (msg.ts != null && newestTs != null && msg.ts <= newestTs) return;  // already loaded via the API
+          if (msg.ts != null && (newestTs == null || msg.ts > newestTs)) newestTs = msg.ts;
+          var box = document.getElementById('messages');
+          var placeholder = box.querySelector('.empty');
+          if (placeholder) placeholder.remove();
+          box.insertAdjacentHTML('afterbegin', logRowHtml(msg));
+          updateCount();
+        };
+        ws_hbd.onclose = function () {
+          setWsStatus(false);
+          setTimeout(WS_Connect, 3000);
+        };
+      }
+
+      // ---- boot ----------------------------------------------------------------
+      fetchLog(true);
+      WS_Connect();
+    </script>
+  </body>
+</html>

hbd/server/templates/nav.html

@@ -6,11 +6,18 @@
     <a href="/live"{% if active_page == "live" %} class="active"{% endif %}>Live Dashboard</a>
     <a href="/plugins"{% if active_page == "plugins" %} class="active"{% endif %}>Host Overview</a>
     <a href="/alerts"{% if active_page == "alerts" %} class="active"{% endif %}>Alerts</a>
-    {% if current_user and current_user.admin %}
+    <a href="/log"{% if active_page == "log" %} class="active"{% endif %}>Log</a>
+    {% if current_user %}
     <a href="/settings"{% if active_page == "settings" %} class="active"{% endif %}>Settings</a>
     {% endif %}
     <a href="/about"{% if active_page == "about" %} class="active"{% endif %}>About</a>
   </div>
+  {% if current_user %}
+  <button id="nav-publish-btn" class="nav-publish-btn" onclick="navPublishConfig()" style="display:none" title="Publish pending config changes to .hb.yaml">&#9888; Publish Config</button>
+  {% endif %}
+  <div class="nav-pie" title="Host alert status">
+    <canvas id="alert-pie" width="44" height="44"></canvas>
+  </div>
   <div class="nav-clock" title="Click for full-screen clock">
     <canvas id="swiss-clock" width="44" height="44"></canvas>
   </div>
@@ -42,4 +49,87 @@
       });
     }
   })();
+
+  function drawAlertPie(critical, warning, ok) {
+    var canvas = document.getElementById('alert-pie');
+    if (!canvas) return;
+    var ctx = canvas.getContext('2d');
+    var SIZE = canvas.width;
+    var R = SIZE / 2;
+    ctx.clearRect(0, 0, SIZE, SIZE);
+    var total = critical + warning + ok;
+    if (total === 0) {
+      ctx.beginPath();
+      ctx.arc(R, R, R - 1, 0, Math.PI * 2);
+      ctx.fillStyle = '#ccc';
+      ctx.fill();
+      return;
+    }
+    var slices = [
+      { value: critical, color: '#e53935' },
+      { value: warning,  color: '#ffb300' },
+      { value: ok,       color: '#43a047' }
+    ];
+    var start = -Math.PI / 2;
+    slices.forEach(function(s) {
+      if (s.value === 0) return;
+      var sweep = (s.value / total) * Math.PI * 2;
+      ctx.beginPath();
+      ctx.moveTo(R, R);
+      ctx.arc(R, R, R - 1, start, start + sweep);
+      ctx.closePath();
+      ctx.fillStyle = s.color;
+      ctx.fill();
+      start += sweep;
+    });
+  }
+
+  function updateAlertPie() {
+    fetch('/api/0/alert_summary').then(function(r) {
+      if (!r.ok) return;
+      return r.json();
+    }).then(function(d) {
+      if (d) drawAlertPie(d.critical || 0, d.warning || 0, d.ok || 0);
+    }).catch(function() {});
+  }
+
+  document.addEventListener('DOMContentLoaded', function() {
+    updateAlertPie();
+    setInterval(updateAlertPie, 30000);
+    navCheckPendingConfig();
+    window.addEventListener('storage', navCheckPendingConfig);
+  });
+
+  function navCheckPendingConfig() {
+    var btn = document.getElementById('nav-publish-btn');
+    if (!btn) return;
+    btn.style.display = localStorage.getItem('hbd_pending_config') ? '' : 'none';
+  }
+
+  async function navPublishConfig() {
+    var btn = document.getElementById('nav-publish-btn');
+    var pending = localStorage.getItem('hbd_pending_config');
+    if (!pending) return;
+    var staged;
+    try { staged = JSON.parse(pending); } catch(e) { return; }
+    if (btn) { btn.disabled = true; btn.textContent = 'Saving…'; }
+    try {
+      var resp = await fetch('/api/0/config', {
+        method: 'POST',
+        headers: {'Content-Type': 'application/json'},
+        body: pending
+      });
+      if (resp.ok) {
+        localStorage.removeItem('hbd_pending_config');
+        window.location.reload();
+      } else {
+        var err = await resp.json().catch(function() { return {}; });
+        alert('Error: ' + (err.error || resp.statusText));
+        if (btn) { btn.disabled = false; btn.textContent = '⚠ Publish Config'; }
+      }
+    } catch(e) {
+      alert('Network error: ' + e.message);
+      if (btn) { btn.disabled = false; btn.textContent = '⚠ Publish Config'; }
+    }
+  }
 </script>

hbd/server/templates/profile.html

@@ -96,7 +96,7 @@
       border-radius: 4px;
       background: #f44336;
       color: #fff;
-      font-size: 0.85em;
+      font-size: 1.00em;
       font-weight: 500;
       text-decoration: none;
       transition: background 0.15s;
@@ -157,7 +157,7 @@
       gap: 6px;
       padding: 4px 12px;
       border-radius: 16px;
-      font-size: 0.85em;
+      font-size: 1.00em;
       font-weight: 500;
       text-decoration: none;
     }
@@ -204,6 +204,119 @@
     }
 
     .channel-name { color: #333; }
+
+    .edit-section { margin-top: 20px; }
+    .edit-section h4 { font-size: .88em; font-weight: 600; color: #333; margin: 0 0 10px; text-transform: uppercase; letter-spacing: .04em; border-bottom: 1px solid #eee; padding-bottom: 6px; }
+    .edit-field { margin-bottom: 10px; }
+    .edit-field label { display: block; font-size: .82em; color: #666; margin-bottom: 3px; }
+    .edit-input { width: 100%; border: 1px solid #ccc; border-radius: 4px; padding: 5px 8px; font-size: .88em; box-sizing: border-box; }
+    .edit-input:focus { border-color: #0066cc; outline: none; }
+    .status-msg { font-size: .82em; margin-left: 8px; }
+    .save-row { display: flex; align-items: center; margin-top: 8px; }
+    .btn-save { background: #0066cc; color: #fff; border: none; border-radius: 4px; padding: 5px 14px; font-size: .85em; cursor: pointer; }
+    .btn-save:hover { background: #0055aa; }
+    /* ---- Channel chip picker ---- */
+    .ch-picker { }
+    .ch-picker-label { font-size: .8em; font-weight: 600; color: #888; text-transform: uppercase; letter-spacing: .04em; margin-bottom: 6px; }
+    .ch-chips { display: flex; flex-wrap: wrap; gap: 6px; min-height: 32px; margin-bottom: 10px; }
+    .ch-chip {
+      display: inline-flex; align-items: center; gap: 5px;
+      padding: 4px 10px; border-radius: 14px; font-size: .85em; font-weight: 500; cursor: pointer;
+      border: none; font-family: inherit;
+    }
+    .ch-chip.selected { background: #e3f2fd; color: #1565c0; }
+    .ch-chip.selected:hover { background: #bbdefb; }
+    .ch-chip.available { background: #f1f3f4; color: #555; }
+    .ch-chip.available:hover { background: #e8eaf6; color: #283593; }
+    .ch-chip-x { font-size: .9em; line-height: 1; color: inherit; opacity: .7; }
+
+    /* ---- My Channels card list ---- */
+    .my-ch-card {
+      border: 1px solid #e8eaf6; border-radius: 6px; margin-bottom: 8px; overflow: hidden;
+    }
+    .my-ch-header {
+      display: flex; align-items: center; gap: 8px; padding: 8px 12px;
+      background: #f8f9ff; border-bottom: 1px solid #e8eaf6;
+    }
+    .my-ch-name { font-weight: 600; font-size: .9em; color: #222; }
+    .my-ch-type { padding: 2px 7px; border-radius: 8px; font-size: .72em; font-weight: 600; background: #e8eaf6; color: #3949ab; }
+    .my-ch-actions { margin-left: auto; display: flex; gap: 5px; }
+    .btn-sm-edit { background: #888; color: #fff; border: none; border-radius: 4px; padding: 2px 8px; font-size: .78em; cursor: pointer; }
+    .btn-sm-edit:hover { background: #666; }
+    .btn-sm-del { background: transparent; color: #c62828; border: 1px solid #e0e0e0; border-radius: 4px; padding: 2px 7px; font-size: .78em; cursor: pointer; }
+    .btn-sm-del:hover { background: #fce4ec; }
+
+    /* ---- Theme picker ---- */
+    .theme-btns { display: flex; gap: 6px; }
+    .theme-btn {
+      padding: 5px 14px;
+      border: 1px solid var(--border, #e0e0e0);
+      border-radius: 4px;
+      background: var(--surface-3, #f5f5f5);
+      color: var(--text-sec, #666);
+      cursor: pointer;
+      font-size: .88em;
+      font-family: inherit;
+    }
+    .theme-btn:hover { border-color: var(--link, #0066cc); color: var(--link, #0066cc); }
+    .theme-btn.active { background: var(--link, #0066cc); color: #fff; border-color: var(--link, #0066cc); }
+
+    /* ── Dark mode ── */
+    html[data-theme="dark"] h1 { color: var(--text); }
+    html[data-theme="dark"] .subtitle { color: var(--text-sec); }
+    html[data-theme="dark"] .profile-card { background: var(--surface); box-shadow: 0 1px 6px var(--shadow); }
+    html[data-theme="dark"] .profile-name { color: var(--text); }
+    html[data-theme="dark"] .profile-username { color: var(--text-sec); }
+    html[data-theme="dark"] .badge-admin { background: #1a3255; color: #7aa8f0; }
+    html[data-theme="dark"] .badge-user  { background: var(--surface-3); color: var(--text-sec); }
+    html[data-theme="dark"] .section { background: var(--surface); box-shadow: 0 1px 6px var(--shadow); }
+    html[data-theme="dark"] .section h2 { color: var(--text); border-bottom-color: var(--border); }
+    html[data-theme="dark"] .settings-row { border-bottom-color: var(--border-4); }
+    html[data-theme="dark"] .settings-label { color: var(--text-sec); }
+    html[data-theme="dark"] .settings-value { color: var(--text); }
+    html[data-theme="dark"] .settings-empty { color: var(--text-dim); }
+    html[data-theme="dark"] .edit-section h4 { color: var(--text); border-bottom-color: var(--border); }
+    html[data-theme="dark"] .edit-field label { color: var(--text-sec); }
+    html[data-theme="dark"] .edit-input { background: var(--input-bg); border-color: var(--input-border); color: var(--text); }
+    html[data-theme="dark"] .channel-row { border-bottom-color: var(--border-4); }
+    html[data-theme="dark"] .channel-name { color: var(--text); }
+    html[data-theme="dark"] .ch-picker-label { color: var(--text-sec); }
+    html[data-theme="dark"] .ch-chip.selected { background: #1a3255; color: #60a5fa; }
+    html[data-theme="dark"] .ch-chip.available { background: var(--surface-3); color: var(--text-sec); }
+    html[data-theme="dark"] .ch-chip.available:hover { background: var(--border); color: var(--link); }
+    html[data-theme="dark"] .my-ch-card { border-color: var(--border); }
+    html[data-theme="dark"] .my-ch-header { background: var(--surface-2); border-bottom-color: var(--border); }
+    html[data-theme="dark"] .my-ch-name { color: var(--text); }
+    html[data-theme="dark"] .host-chip.owner   { background: #0d2e17; color: #66bb6a; }
+    html[data-theme="dark"] .host-chip.manager { background: #0d1f40; color: #64b5f6; }
+    html[data-theme="dark"] .host-chip.monitor { background: #1e0d30; color: #ba68c8; }
+    html[data-theme="dark"] .no-hosts { color: var(--text-dim); }
+    html[data-theme="dark"] .ch-modal-box { background: var(--surface); color: var(--text); }
+    html[data-theme="dark"] .ch-modal-box h3 { color: var(--text); }
+    html[data-theme="dark"] .ch-form-row label { color: var(--text-sec); }
+    html[data-theme="dark"] .ch-form-divider { color: var(--text-muted); border-top-color: var(--border); }
+
+    /* ---- Channel modal (for My Channels CRUD) ---- */
+    .ch-modal-overlay {
+      position: fixed; inset: 0; background: rgba(0,0,0,.4);
+      display: flex; align-items: center; justify-content: center; z-index: 1001;
+    }
+    .ch-modal-box {
+      background: #fff; border-radius: 8px; padding: 24px;
+      min-width: 360px; max-width: 520px; width: 95%;
+      box-shadow: 0 8px 32px rgba(0,0,0,.2);
+    }
+    .ch-modal-box h3 { margin: 0 0 16px; font-size: 1em; }
+    .ch-form-row { margin-bottom: 12px; }
+    .ch-form-row label { display: block; font-size: .83em; font-weight: 600; color: #555; margin-bottom: 3px; }
+    .ch-form-row input[type=text], .ch-form-row input[type=password], .ch-form-row select {
+      width: 100%; border: 1px solid #ccc; border-radius: 4px; padding: 5px 8px;
+      font-size: .88em; box-sizing: border-box; font-family: inherit;
+    }
+    .ch-form-row input:focus, .ch-form-row select:focus { border-color: #0066cc; outline: none; }
+    .ch-form-divider { font-size: .78em; font-weight: 700; text-transform: uppercase; letter-spacing: .05em; color: #888; margin: 14px 0 8px; border-top: 1px solid #eee; padding-top: 10px; }
+    .ch-modal-footer { display: flex; justify-content: flex-end; gap: 8px; margin-top: 18px; }
+    .ch-modal-status { font-size: .83em; margin-top: 8px; }
   </style>
 
   <body>
@@ -266,19 +379,158 @@
         </div>
       </div>
 
-      <!-- Notification channels -->
+      {% if current_user %}
+      <!-- ---- Editable identity ---- -->
+      <div class="section edit-section">
+        <h4>Identity</h4>
+        <div class="edit-field">
+          <label for="profile-fullname">Display name</label>
+          <input id="profile-fullname" class="edit-input" type="text" value="{{ current_user.full_name | e }}" placeholder="Full name">
+        </div>
+        <div class="edit-field">
+          <label for="profile-avatar">Avatar URL or path</label>
+          <input id="profile-avatar" class="edit-input" type="text" value="{{ current_user.avatar | e }}" placeholder="/path/to/avatar.png or https://…">
+        </div>
+        <div class="save-row">
+          <button class="btn-save" onclick="saveIdentity()">Save</button>
+          <span id="identity-status" class="status-msg"></span>
+        </div>
+      </div>
+
+      <!-- ---- Change password ---- -->
+      <div class="section edit-section">
+        <h4>Change password</h4>
+        <div class="edit-field">
+          <label for="profile-current-pw">Current password</label>
+          <input id="profile-current-pw" class="edit-input" type="password" autocomplete="current-password">
+        </div>
+        <div class="edit-field">
+          <label for="profile-new-pw">New password</label>
+          <input id="profile-new-pw" class="edit-input" type="password" autocomplete="new-password">
+        </div>
+        <div class="save-row">
+          <button class="btn-save" onclick="changePassword()">Change password</button>
+          <span id="password-status" class="status-msg"></span>
+        </div>
+      </div>
+      {% endif %}
+
+      <!-- Notification channels — chip picker -->
       <div class="section">
         <h2>Notification Channels</h2>
-        {% if notification_channels %}
+        {% if current_user %}
-        {% for ch in notification_channels %}
+        <p style="font-size:.82em;color:#888;margin:0 0 12px">Click a channel to add or remove it from your alert list.</p>
-        <div class="channel-row">
+        {% if all_channels %}
-          <span class="channel-type">{{ ch.type }}</span>
+        <div class="ch-picker">
-          <span class="channel-name">{{ ch.name }}</span>
+          <div class="ch-picker-label">Selected</div>
+          <div id="selected-chips" class="ch-chips">
+            {% for ch in all_channels %}
+            {% if ch.name in (current_user.notification_channels or []) %}
+            <button class="ch-chip selected" data-ch="{{ ch.name | e }}" onclick="toggleChip(this)">
+              {{ ch.name | e }} <span class="ch-chip-x">×</span>
+            </button>
+            {% endif %}
+            {% endfor %}
+            {% set selected_set = current_user.notification_channels or [] %}
+            {% set has_selected = selected_set | length > 0 %}
+            {% if not has_selected %}
+            <span style="font-size:.83em;color:#bbb;font-style:italic;align-self:center">None selected</span>
+            {% endif %}
+          </div>
+          <div class="ch-picker-label">Available</div>
+          <div id="available-chips" class="ch-chips">
+            {% for ch in all_channels %}
+            {% if ch.name not in (current_user.notification_channels or []) %}
+            <button class="ch-chip available" data-ch="{{ ch.name | e }}" onclick="toggleChip(this)">
+              + {{ ch.name | e }}
+            </button>
+            {% endif %}
+            {% endfor %}
+          </div>
         </div>
-        {% endfor %}
         {% else %}
-        <span class="no-hosts">No personal notification channels configured.</span>
+        <p style="font-size:.83em;color:#bbb;font-style:italic">No notification channels available. You can create your own below.</p>
         {% endif %}
+        <div class="save-row">
+          <button class="btn-save" onclick="saveChannels()">Save channels</button>
+          <span id="channels-status" class="status-msg"></span>
+        </div>
+        {% else %}
+        <span class="no-hosts">Log in to manage notification channels.</span>
+        {% endif %}
+      </div>
+
+      <!-- My Channels — create/edit/delete own channels -->
+      {% if current_user %}
+      <div class="section">
+        <h2>My Channels</h2>
+        <p style="font-size:.82em;color:#888;margin:0 0 12px">Channels you own are private to you. Global channels are managed by administrators.</p>
+        <div id="my-channels-list">
+          {% set my_channels = all_channels | selectattr('owner', 'equalto', current_user.username) | list %}
+          {% for ch in my_channels %}
+          <div class="my-ch-card" id="mychcard-{{ ch.name | e }}">
+            <div class="my-ch-header">
+              <span class="my-ch-name">{{ ch.name | e }}</span>
+              <span class="my-ch-type">{{ ch.type | e }}</span>
+              <span class="my-ch-actions">
+                <button class="btn-sm-edit" onclick="openMyChModal('{{ ch.name | e }}')">Edit</button>
+                <button class="btn-sm-del" onclick="deleteMyChannel('{{ ch.name | e }}')">✕</button>
+              </span>
+            </div>
+          </div>
+          {% endfor %}
+          {% if not my_channels %}
+          <p id="my-channels-empty" style="font-size:.83em;color:#bbb;font-style:italic">No channels yet.</p>
+          {% endif %}
+        </div>
+        <div class="save-row" style="margin-top:8px">
+          <button class="btn-save" onclick="openMyChModal()">+ New channel</button>
+        </div>
+      </div>
+
+      <!-- My Channels modal -->
+      <div id="my-ch-modal" class="ch-modal-overlay" style="display:none" onclick="if(event.target===this)closeMyChModal()">
+        <div class="ch-modal-box">
+          <h3 id="my-ch-modal-title">New Channel</h3>
+          <div class="ch-form-row">
+            <label>Channel name</label>
+            <input type="text" id="my-ch-name" placeholder="e.g. my_pushover" autocomplete="off">
+          </div>
+          <div class="ch-form-row">
+            <label>Type</label>
+            <select id="my-ch-type" onchange="onMyChTypeChange()">
+              <option value="">— select —</option>
+            </select>
+          </div>
+          <div id="my-ch-type-fields"></div>
+          <div class="ch-form-divider">Options</div>
+          <div class="ch-form-row">
+            <label>Minimum alert level</label>
+            <select id="my-ch-min-level">
+              <option value="WARNING">WARNING (and above)</option>
+              <option value="CRITICAL">CRITICAL only</option>
+            </select>
+          </div>
+          <div id="my-ch-modal-status" class="ch-modal-status"></div>
+          <div class="ch-modal-footer">
+            <button class="btn-save" style="background:#888" onclick="closeMyChModal()">Cancel</button>
+            <button class="btn-save" onclick="saveMyChannel()">Save</button>
+          </div>
+        </div>
+      </div>
+      {% endif %}
+
+      <!-- Appearance -->
+      <div class="section">
+        <h2>Appearance</h2>
+        <div class="settings-row">
+          <span class="settings-label">Theme</span>
+          <div class="theme-btns">
+            <button class="theme-btn" data-theme-val="auto"  onclick="setTheme('auto')">Auto</button>
+            <button class="theme-btn" data-theme-val="light" onclick="setTheme('light')">Light</button>
+            <button class="theme-btn" data-theme-val="dark"  onclick="setTheme('dark')">Dark</button>
+          </div>
+        </div>
       </div>
 
       <!-- Host access -->
@@ -326,5 +578,255 @@
       </div>
 
     </div>
+    <script>
+      // ---- Theme ----
+      function applyTheme(pref) {
+        var dark = pref === 'dark' ||
+          (pref === 'auto' && window.matchMedia('(prefers-color-scheme: dark)').matches);
+        if (dark) { document.documentElement.setAttribute('data-theme', 'dark'); }
+        else { document.documentElement.removeAttribute('data-theme'); }
+      }
+      function setTheme(pref) {
+        try { localStorage.setItem('hbd_theme', pref); } catch(e) {}
+        applyTheme(pref);
+        document.querySelectorAll('.theme-btn').forEach(function(b) {
+          b.classList.toggle('active', b.dataset.themeVal === pref);
+        });
+      }
+      (function() {
+        var pref = 'auto';
+        try { pref = localStorage.getItem('hbd_theme') || 'auto'; } catch(e) {}
+        document.querySelectorAll('.theme-btn').forEach(function(b) {
+          b.classList.toggle('active', b.dataset.themeVal === pref);
+        });
+      })();
+
+      // ---- Identity ----
+      async function saveIdentity() {
+        const full_name = document.getElementById('profile-fullname').value;
+        const avatar = document.getElementById('profile-avatar').value;
+        const resp = await fetch('/api/0/users/me', {
+          method: 'PUT',
+          headers: {'Content-Type': 'application/json'},
+          body: JSON.stringify({full_name, avatar}),
+        });
+        if (resp.ok) {
+          showStatus('identity-status', 'Saved', '#2e7d32');
+        } else {
+          const err = await resp.json().catch(() => ({}));
+          showStatus('identity-status', err.error || 'Error saving', '#c62828');
+        }
+      }
+
+      // ---- Password ----
+      async function changePassword() {
+        const current = document.getElementById('profile-current-pw').value;
+        const newpw = document.getElementById('profile-new-pw').value;
+        if (!current || !newpw) {
+          showStatus('password-status', 'Both fields are required', '#c62828');
+          return;
+        }
+        const resp = await fetch('/api/0/users/me', {
+          method: 'PUT',
+          headers: {'Content-Type': 'application/json'},
+          body: JSON.stringify({password: {current, new: newpw}}),
+        });
+        if (resp.ok) {
+          document.getElementById('profile-current-pw').value = '';
+          document.getElementById('profile-new-pw').value = '';
+          showStatus('password-status', 'Password changed', '#2e7d32');
+        } else {
+          const err = await resp.json().catch(() => ({}));
+          showStatus('password-status', err.error || 'Error', '#c62828');
+        }
+      }
+
+      // ---- Channel chip picker ----
+      function toggleChip(btn) {
+        const name = btn.dataset.ch;
+        const isSelected = btn.classList.contains('selected');
+        if (isSelected) {
+          // Move to available
+          btn.classList.remove('selected');
+          btn.classList.add('available');
+          btn.innerHTML = '+ ' + escHtml(name);
+          btn.onclick = function() { toggleChip(this); };
+          document.getElementById('available-chips').appendChild(btn);
+          // Remove "None selected" placeholder if it exists
+        } else {
+          // Move to selected
+          btn.classList.remove('available');
+          btn.classList.add('selected');
+          btn.innerHTML = escHtml(name) + ' <span class="ch-chip-x">×</span>';
+          btn.onclick = function() { toggleChip(this); };
+          document.getElementById('selected-chips').appendChild(btn);
+        }
+        // Update placeholder visibility
+        const sel = document.getElementById('selected-chips');
+        const placeholder = sel.querySelector('span[style]');
+        const hasChips = sel.querySelectorAll('.ch-chip.selected').length > 0;
+        if (placeholder) placeholder.style.display = hasChips ? 'none' : '';
+      }
+
+      async function saveChannels() {
+        const notification_channels = [
+          ...document.querySelectorAll('#selected-chips .ch-chip.selected')
+        ].map(b => b.dataset.ch);
+        const resp = await fetch('/api/0/users/me', {
+          method: 'PUT',
+          headers: {'Content-Type': 'application/json'},
+          body: JSON.stringify({notification_channels}),
+        });
+        if (resp.ok) {
+          showStatus('channels-status', 'Saved', '#2e7d32');
+        } else {
+          const err = await resp.json().catch(() => ({}));
+          showStatus('channels-status', err.error || 'Error saving', '#c62828');
+        }
+      }
+
+      // ---- My Channels CRUD ----
+      let _myChSchemas = {};
+      let _myChEditName = null;
+
+      async function _loadMyChSchemas() {
+        try {
+          const r = await fetch('/api/0/notification_channel_types');
+          _myChSchemas = await r.json();
+          const sel = document.getElementById('my-ch-type');
+          if (!sel) return;
+          Object.entries(_myChSchemas).forEach(([k, v]) => {
+            const opt = document.createElement('option');
+            opt.value = k; opt.textContent = v.label;
+            sel.appendChild(opt);
+          });
+        } catch(e) { console.warn('Could not load channel schemas', e); }
+      }
+
+      function onMyChTypeChange() {
+        const type = document.getElementById('my-ch-type').value;
+        const container = document.getElementById('my-ch-type-fields');
+        container.innerHTML = '';
+        if (!type || !_myChSchemas[type]) return;
+        const divider = document.createElement('div');
+        divider.className = 'ch-form-divider';
+        divider.textContent = _myChSchemas[type].label + ' settings';
+        container.appendChild(divider);
+        (_myChSchemas[type].fields || []).forEach(sf => {
+          const row = document.createElement('div');
+          row.className = 'ch-form-row';
+          const lbl = document.createElement('label');
+          lbl.textContent = sf.label + (sf.required ? ' *' : '');
+          const inp = document.createElement('input');
+          inp.type = sf.type === 'secret' ? 'password' : 'text';
+          inp.id = 'mychf-' + sf.key;
+          inp.placeholder = sf.required ? '(required)' : '(optional)';
+          inp.autocomplete = 'off';
+          row.appendChild(lbl);
+          row.appendChild(inp);
+          container.appendChild(row);
+        });
+      }
+
+      async function openMyChModal(name) {
+        _myChEditName = name || null;
+        document.getElementById('my-ch-modal-status').textContent = '';
+        document.getElementById('my-ch-modal-title').textContent = name ? 'Edit Channel' : 'New Channel';
+        document.getElementById('my-ch-name').value = name || '';
+        document.getElementById('my-ch-name').disabled = !!name;
+        document.getElementById('my-ch-type').value = '';
+        document.getElementById('my-ch-type-fields').innerHTML = '';
+        document.getElementById('my-ch-min-level').value = 'WARNING';
+
+        if (name) {
+          try {
+            const r = await fetch('/api/0/notification_channels');
+            const channels = await r.json();
+            const ch = channels.find(c => c.name === name);
+            if (ch) {
+              document.getElementById('my-ch-type').value = ch.type;
+              onMyChTypeChange();
+              document.getElementById('my-ch-min-level').value = ch.min_level || 'WARNING';
+              (ch.fields || []).forEach(f => {
+                const inp = document.getElementById('mychf-' + f.key);
+                if (inp) inp.value = f.value || '';
+              });
+            }
+          } catch(e) { console.warn('Failed to load channel', e); }
+        }
+        document.getElementById('my-ch-modal').style.display = 'flex';
+      }
+
+      function closeMyChModal() {
+        document.getElementById('my-ch-modal').style.display = 'none';
+      }
+
+      async function saveMyChannel() {
+        const name = document.getElementById('my-ch-name').value.trim();
+        const type = document.getElementById('my-ch-type').value;
+        const minLevel = document.getElementById('my-ch-min-level').value;
+        const statusEl = document.getElementById('my-ch-modal-status');
+        statusEl.textContent = '';
+
+        if (!name) { statusEl.textContent = 'Name is required.'; statusEl.style.color = '#c62828'; return; }
+        if (!type) { statusEl.textContent = 'Please select a type.'; statusEl.style.color = '#c62828'; return; }
+
+        const body = { name, type, min_level: minLevel };
+        if (_myChSchemas[type]) {
+          (_myChSchemas[type].fields || []).forEach(sf => {
+            const inp = document.getElementById('mychf-' + sf.key);
+            if (inp) body[sf.key] = inp.value;
+          });
+        }
+
+        const isEdit = !!_myChEditName;
+        const url = isEdit
+          ? '/api/0/notification_channels/' + encodeURIComponent(_myChEditName)
+          : '/api/0/notification_channels';
+        const method = isEdit ? 'PUT' : 'POST';
+        try {
+          const r = await fetch(url, { method, headers: {'Content-Type': 'application/json'}, body: JSON.stringify(body) });
+          if (r.ok) {
+            closeMyChModal();
+            window.location.reload();
+          } else {
+            const err = await r.json().catch(() => ({}));
+            statusEl.textContent = err.error || 'Error saving.';
+            statusEl.style.color = '#c62828';
+          }
+        } catch(e) {
+          statusEl.textContent = 'Network error: ' + e.message;
+          statusEl.style.color = '#c62828';
+        }
+      }
+
+      async function deleteMyChannel(name) {
+        if (!confirm('Delete channel "' + name + '"?')) return;
+        try {
+          const r = await fetch('/api/0/notification_channels/' + encodeURIComponent(name), { method: 'DELETE' });
+          if (r.ok) {
+            window.location.reload();
+          } else {
+            const err = await r.json().catch(() => ({}));
+            alert('Error: ' + (err.error || 'Could not delete.'));
+          }
+        } catch(e) { alert('Network error: ' + e.message); }
+      }
+
+      // ---- Utilities ----
+      function showStatus(id, msg, color) {
+        const el = document.getElementById(id);
+        if (!el) return;
+        el.textContent = msg;
+        el.style.color = color;
+        setTimeout(() => { el.textContent = ''; }, 3000);
+      }
+
+      function escHtml(s) {
+        return String(s).replace(/&/g,'&amp;').replace(/</g,'&lt;').replace(/>/g,'&gt;').replace(/"/g,'&quot;');
+      }
+
+      document.addEventListener('DOMContentLoaded', _loadMyChSchemas);
+    </script>
   </body>
 </html>

hbd/server/threshold.py

@@ -9,10 +9,11 @@ This module provides a flexible threshold checking system that:
 - Supports multiple comparison operators
 """
 
+import asyncio
 import logging
 import time
 from enum import Enum
-from typing import Dict, Any, Optional, Tuple, Callable
+from typing import Dict, List, Any, Optional, Tuple, Callable
 from . import notify as notify_mod
 from .config import THRESHOLD_DEFAULTS
 
@@ -29,12 +30,13 @@ class AlertLevel(Enum):
 
 class ComparisonOperator(Enum):
     """Supported comparison operators for threshold checks."""
-    GT = ">"      # Greater than
+    GT = ">"        # Greater than
-    GTE = ">="    # Greater than or equal
+    GTE = ">="      # Greater than or equal
-    LT = "<"      # Less than
+    LT = "<"        # Less than
-    LTE = "<="    # Less than or equal
+    LTE = "<="      # Less than or equal
-    EQ = "=="     # Equal to
+    EQ = "=="       # Equal to
-    NEQ = "!="    # Not equal to
+    NEQ = "!="      # Not equal to
+    NAGIOS = "nagios"  # Nagios exit-code semantics: 0=OK 1=WARNING 2=CRITICAL 3=UNKNOWN
 
 
 class AlertState:
@@ -56,6 +58,7 @@ class AlertState:
         self.last_notification = None
         self.threshold_value = None  # The threshold value that triggered alert
         self.operator = None  # The comparison operator (>, <, >=, etc.)
+        self.hysteresis: Optional[float] = None  # Hysteresis fraction used for recovery
         self.formatted_message = None  # Formatted display message for UI
         self.acknowledged = False  # Whether alert has been acknowledged
         self.acknowledged_at = None  # Timestamp when acknowledged
@@ -151,6 +154,15 @@ class AlertState:
         if self.formatted_message is not None:
             result["formatted_message"] = self.formatted_message
 
+        # Compute and expose the recovery threshold so the UI can display it
+        if (self.hysteresis and self.threshold_value is not None
+                and self.operator is not None):
+            ha = abs(self.threshold_value * self.hysteresis)
+            if self.operator in ('>', '>='):
+                result["recovery_threshold"] = round(self.threshold_value - ha, 4)
+            elif self.operator in ('<', '<='):
+                result["recovery_threshold"] = round(self.threshold_value + ha, 4)
+
         return result
     
     def __setstate__(self, state):
@@ -158,6 +170,8 @@ class AlertState:
         self.__dict__.update(state)
         if not hasattr(self, 'consecutive_count'):
             self.consecutive_count = 0
+        if not hasattr(self, 'hysteresis'):
+            self.hysteresis = None
 
     def acknowledge(self):
         """Acknowledge this alert to stop reminder notifications."""
@@ -181,6 +195,7 @@ class ThresholdConfig:
         hysteresis: float = 0.0,
         enabled: bool = True,
         count: int = 1,
+        grace: Optional[float] = None,
     ):
         """
         Initialize threshold configuration.
@@ -193,6 +208,7 @@ class ThresholdConfig:
             hysteresis: Hysteresis percentage to prevent flapping (0.0-1.0)
             enabled: Whether this threshold is enabled
             count: Number of consecutive exceedances required before alerting (default 1)
+            grace: Per-metric grace period in seconds; overrides global grace when set
         """
         self.metric_path = metric_path
         self.warning = warning
@@ -201,6 +217,7 @@ class ThresholdConfig:
         self.hysteresis = hysteresis
         self.display = display
         self.count = max(1, int(count))
+        self.grace = float(grace) if grace is not None else None
         
         # Parse operator
         try:
@@ -226,6 +243,16 @@ class ThresholdConfig:
         if not self.enabled:
             return AlertLevel.OK
 
+        # Nagios exit-code semantics: value IS the severity
+        if self.operator == ComparisonOperator.NAGIOS:
+            try:
+                code = int(value)
+            except (TypeError, ValueError):
+                return AlertLevel.UNKNOWN
+            return {0: AlertLevel.OK, 1: AlertLevel.WARNING, 2: AlertLevel.CRITICAL}.get(
+                code, AlertLevel.UNKNOWN
+            )
+
         try:
             # Convert value to float for comparison
             value = float(value)
@@ -262,6 +289,10 @@ class ThresholdConfig:
         """
         new_level = self.evaluate(value)
 
+        # Nagios exit codes are discrete integers — hysteresis doesn't apply
+        if self.operator == ComparisonOperator.NAGIOS:
+            return new_level
+
         # If no hysteresis, return new level
         if self.hysteresis == 0.0:
             return new_level
@@ -328,14 +359,17 @@ class ThresholdChecker:
             renotify_interval: Seconds between repeat notifications (default: 1 hour)
             journal: Optional MessageJournal instance for logging threshold events
         """
-        # Named threshold configurations: {config_name: {metric_path: ThresholdConfig}}
+        # Named threshold configurations (pre-merged: defaults + overrides): {config_name: {metric_path: ThresholdConfig}}
         self.threshold_configs = {}
 
+        # Raw overrides only for each named config (no defaults baked in): {config_name: {metric_path: ThresholdConfig}}
+        self.threshold_raw_configs: Dict[str, Dict[str, ThresholdConfig]] = {}
+
         # Single threshold set for backward compatibility: {metric_path: ThresholdConfig}
         self.thresholds = {}
 
-        # Host to config name mapping: {host_name: config_name}
+        # Host to ordered list of config names: {host_name: [config_name, ...]}
-        self.host_config_mapping = {}
+        self.host_config_mapping: Dict[str, List[str]] = {}
 
         # Default config name to use when no mapping exists
         self.default_config = "default"
@@ -372,6 +406,7 @@ class ThresholdChecker:
         
         # Clear old configuration
         self.threshold_configs.clear()
+        self.threshold_raw_configs.clear()
         self.thresholds.clear()
         self.host_config_mapping.clear()
         self.grace_seconds = float(config.get("grace", 2))
@@ -391,10 +426,24 @@ class ThresholdChecker:
         Supports two formats:
         1. Legacy format with direct 'thresholds' section
         2. New format with 'threshold_configs' and 'host_threshold_mapping'
+
+        In all cases, THRESHOLD_DEFAULTS are seeded into threshold_configs["default"]
+        so the Settings page always shows the built-in defaults.
+        _parse_multi_config() overwrites this with the fully-merged effective defaults.
         """
+        # Always expose built-in defaults through threshold_configs["default"] so
+        # the Settings page has something to display even in legacy/no-config mode.
+        seed: Dict[str, ThresholdConfig] = {}
+        for plugin_name, plugin_thresholds in THRESHOLD_DEFAULTS.get("thresholds", {}).items():
+            if isinstance(plugin_thresholds, dict):
+                self._parse_plugin_thresholds(plugin_name, plugin_thresholds, target_dict=seed)
+        if seed:
+            self.threshold_configs["default"] = seed
+            self.threshold_raw_configs["default"] = {}
+
         # Check for new multi-config format
         if "threshold_configs" in config:
-            self._parse_multi_config(config)
+            self._parse_multi_config(config)  # overwrites threshold_configs["default"]
         elif "thresholds" in config:
             # Legacy single threshold configuration
             self._parse_legacy_config(config)
@@ -424,9 +473,10 @@ class ThresholdChecker:
                         self._parse_plugin_thresholds(plugin_name, plugin_thresholds, target_dict=effective_defaults)
 
         self.threshold_configs["default"] = dict(effective_defaults)
+        self.threshold_raw_configs["default"] = {}
         logger.info("Registered 'default' threshold config with %d metrics", len(effective_defaults))
 
-        # Parse each named configuration, seeding it with effective_defaults first
+        # Parse each named configuration
         for config_name, config_data in threshold_configs.items():
             if config_name == "default":
                 continue  # already handled above
@@ -440,33 +490,61 @@ class ThresholdChecker:
                 continue
 
             logger.info("Parsing threshold configuration: %s", config_name)
-            self.threshold_configs[config_name] = dict(effective_defaults)
 
+            # Raw overrides only (used for multi-config layering)
+            raw_overrides: Dict[str, ThresholdConfig] = {}
             thresholds_config = config_data["thresholds"]
             for plugin_name, plugin_thresholds in thresholds_config.items():
                 if not isinstance(plugin_thresholds, dict):
                     continue
+                plugin_enabled = plugin_thresholds.get('enabled', plugin_thresholds.get('enable', True))
+                if not plugin_enabled:
+                    # raw_overrides is empty at this point so there's nothing to delete.
+                    # Instead, inject disabled stubs for every matching effective_default so
+                    # the merge step overwrites the inherited defaults.
+                    for key, tc in effective_defaults.items():
+                        if key.startswith(f"{plugin_name}."):
+                            raw_overrides[key] = ThresholdConfig(
+                                metric_path=key,
+                                warning=tc.warning,
+                                critical=tc.critical,
+                                operator=tc.operator.value,
+                                enabled=False,
+                            )
+                    logger.info(
+                        "Plugin-level disable in config '%s': disabled all thresholds for %s",
+                        config_name, plugin_name,
+                    )
+                else:
+                    self._parse_plugin_thresholds(plugin_name, plugin_thresholds, target_dict=raw_overrides)
+            self.threshold_raw_configs[config_name] = raw_overrides
 
-                self._parse_plugin_thresholds(
+            # Pre-merged version (defaults + overrides) for single-config fast path
-                    plugin_name,
+            self.threshold_configs[config_name] = dict(effective_defaults)
-                    plugin_thresholds,
+            self.threshold_configs[config_name].update(raw_overrides)
-                    target_dict=self.threshold_configs[config_name]
-                )
 
-        # Parse host to config mapping from two possible sources
+        # Parse host → config list mapping from two possible sources
-        # 1. New format: hosts section with threshold_config attribute
+
+        def _normalise(value) -> List[str]:
+            """Accept a string or list; always return a list."""
+            if isinstance(value, list):
+                return [str(v) for v in value]
+            return [str(value)]
+
+        # 1. hosts section with threshold_config attribute (string or list)
         if "hosts" in config:
             hosts_config = config["hosts"]
             if isinstance(hosts_config, dict):
                 for host_name, host_attrs in hosts_config.items():
                     if isinstance(host_attrs, dict) and "threshold_config" in host_attrs:
-                        self.host_config_mapping[host_name] = host_attrs["threshold_config"]
+                        self.host_config_mapping[host_name] = _normalise(host_attrs["threshold_config"])
 
-        # 2. Legacy format: host_threshold_mapping section (for backward compatibility)
+        # 2. Legacy host_threshold_mapping section (string values only)
         if "host_threshold_mapping" in config:
             legacy_mapping = config.get("host_threshold_mapping", {})
             if isinstance(legacy_mapping, dict):
-                self.host_config_mapping.update(legacy_mapping)
+                for host_name, value in legacy_mapping.items():
+                    self.host_config_mapping[host_name] = _normalise(value)
         
         # Set default config (first one alphabetically or explicitly set)
         self.default_config = config.get("default_threshold_config", "default")
@@ -516,14 +594,26 @@ class ThresholdChecker:
             self._parse_rtt_thresholds(thresholds, target_dict)
             return
 
+        # Plugin-level enabled: false (also accept 'enable' as a common typo) removes all
+        # thresholds for this plugin — e.g. memory_monitor: {enabled: false}.
+        plugin_enabled = thresholds.get('enabled', thresholds.get('enable', True))
+        if not plugin_enabled:
+            for key in [k for k in target_dict if k.startswith(f"{plugin_name}.")]:
+                del target_dict[key]
+            logger.info("Plugin-level disable: removed all thresholds for %s", plugin_name)
+            return
+
         for metric_name, threshold_config in thresholds.items():
             if not isinstance(threshold_config, dict):
                 continue
             
-            # Handle nested metrics (e.g., partitions./.percent)
+            # Handle nested metrics (e.g., partitions./.percent or pools.*.status)
             if metric_name == "partitions":
                 self._parse_partition_thresholds(plugin_name, threshold_config, target_dict)
                 continue
+            if metric_name == "pools":
+                self._parse_pool_thresholds(plugin_name, threshold_config, target_dict)
+                continue
             
             metric_path = f"{plugin_name}.{metric_name}"
             
@@ -531,11 +621,15 @@ class ThresholdChecker:
             warning = threshold_config.get("warning")
             critical = threshold_config.get("critical")
             operator = threshold_config.get("operator", ">")
-            display = threshold_config.get("display", "(threshold: {op_symbol} {threshold_value})")
+            # Nagios operator maps exit codes directly; no numeric thresholds needed
-            hysteresis = threshold_config.get("hysteresis", 0.1)  # 10% default
+            is_nagios_op = (operator == "nagios")
+            default_display = "{check_name}: {output}" if is_nagios_op else "(threshold: {op_symbol} {threshold_value})"
+            display = threshold_config.get("display", default_display)
+            hysteresis = threshold_config.get("hysteresis", 0.0 if is_nagios_op else 0.02)
             enabled = threshold_config.get("enabled", True)
+            grace = threshold_config.get("grace", None)
 
-            if warning is None and critical is None:
+            if warning is None and critical is None and not is_nagios_op:
                 logger.warning("No thresholds defined for %s, skipping", metric_path)
                 continue
 
@@ -546,7 +640,8 @@ class ThresholdChecker:
                 operator=operator,
                 hysteresis=hysteresis,
                 enabled=enabled,
-                display=display
+                display=display,
+                grace=grace,
             )
             
             target_dict[metric_path] = threshold
@@ -591,6 +686,7 @@ class ThresholdChecker:
                 hysteresis = threshold_config.get("hysteresis", 0.1)
                 enabled = threshold_config.get("enabled", True)
                 display = threshold_config.get("display")
+                grace = threshold_config.get("grace", None)
                 if warning is None and critical is None:
                     continue
 
@@ -601,11 +697,64 @@ class ThresholdChecker:
                     operator=operator,
                     hysteresis=hysteresis,
                     enabled=enabled,
-                    display=display 
+                    display=display,
+                    grace=grace,
                 )
                 
                 target_dict[metric_path] = threshold
 
+    def _parse_pool_thresholds(
+        self,
+        plugin_name: str,
+        pools: Dict[str, Any],
+        target_dict: Optional[Dict[str, ThresholdConfig]] = None,
+    ):
+        """Parse ZFS pool thresholds.  Pool names may be literal or '*' (all pools).
+
+        Config shape::
+
+            zfs_monitor:
+              pools:
+                '*':
+                  status:
+                    warning: 1
+                    critical: 2
+                    operator: '>'
+                tank:
+                  capacity:
+                    warning: 80
+                    critical: 90
+        """
+        if target_dict is None:
+            target_dict = self.thresholds
+
+        for pool_name, metrics in pools.items():
+            if not isinstance(metrics, dict):
+                continue
+            for metric_name, threshold_config in metrics.items():
+                if not isinstance(threshold_config, dict):
+                    continue
+                metric_path = f"{plugin_name}.{pool_name}.{metric_name}"
+                warning = threshold_config.get("warning")
+                critical = threshold_config.get("critical")
+                operator = threshold_config.get("operator", ">")
+                hysteresis = threshold_config.get("hysteresis", 0.02)
+                enabled = threshold_config.get("enabled", True)
+                display = threshold_config.get("display")
+                grace = threshold_config.get("grace", None)
+                if warning is None and critical is None:
+                    continue
+                target_dict[metric_path] = ThresholdConfig(
+                    metric_path=metric_path,
+                    warning=warning,
+                    critical=critical,
+                    operator=operator,
+                    hysteresis=hysteresis,
+                    enabled=enabled,
+                    display=display,
+                    grace=grace,
+                )
+
     def _parse_rtt_thresholds(
         self,
         rtt_thresholds: Dict[str, Any],
@@ -635,10 +784,11 @@ class ThresholdChecker:
         warning = rtt_thresholds.get("warning")
         critical = rtt_thresholds.get("critical")
         operator = rtt_thresholds.get("operator", ">")
-        hysteresis = rtt_thresholds.get("hysteresis", 0.1)  # 10% default
+        hysteresis = rtt_thresholds.get("hysteresis", 0.02)  # 2% default
         enabled = rtt_thresholds.get("enabled", True)
         display = rtt_thresholds.get("display")
         count = rtt_thresholds.get("count", 1)
+        grace = rtt_thresholds.get("grace", None)
 
         if warning is None and critical is None:
             logger.warning("No RTT thresholds defined, skipping")
@@ -653,6 +803,7 @@ class ThresholdChecker:
             enabled=enabled,
             display=display,
             count=count,
+            grace=grace,
         )
 
         target_dict[metric_path] = threshold
@@ -664,7 +815,10 @@ class ThresholdChecker:
         )
     
     def get_thresholds_for_host(self, host_name: str) -> Dict[str, ThresholdConfig]:
-        """Get the appropriate threshold configuration for a host.
+        """Get the effective threshold configuration for a host.
+
+        When threshold_config is a list, configs are applied left-to-right on top
+        of the default thresholds so earlier entries can be overridden by later ones.
 
         Args:
             host_name: Name of the host
@@ -676,23 +830,40 @@ class ThresholdChecker:
         if self.thresholds and not self.threshold_configs:
             return self.thresholds
 
-        # Multi-config mode: look up host-specific configuration
+        if not self.threshold_configs:
-        if self.threshold_configs:
+            return {}
-            config_name = self.host_config_mapping.get(host_name, self.default_config)
 
-            if config_name in self.threshold_configs:
+        config_names = self.host_config_mapping.get(host_name)
-                return self.threshold_configs[config_name]
+
-            else:
+        # No host-specific mapping → return pre-merged default
+        if not config_names:
+            return self.threshold_configs.get(self.default_config, {})
+
+        # Single config → fast path using pre-merged copy
+        if len(config_names) == 1:
+            name = config_names[0]
+            if name in self.threshold_configs:
+                return self.threshold_configs[name]
+            logger.warning(
+                "Threshold config '%s' not found for host '%s', using default '%s'",
+                name, host_name, self.default_config,
+            )
+            return self.threshold_configs.get(self.default_config, {})
+
+        # Multiple configs → start from defaults, layer raw overrides in order
+        result = dict(self.threshold_configs.get(self.default_config, {}))
+        for name in config_names:
+            if name == self.default_config:
+                continue  # defaults already the base
+            raw = self.threshold_raw_configs.get(name)
+            if raw is None:
                 logger.warning(
-                    "Threshold config '%s' not found for host '%s', using default '%s'",
+                    "Threshold config '%s' not found for host '%s', skipping",
-                    config_name,
+                    name, host_name,
-                    host_name,
-                    self.default_config
                 )
-                return self.threshold_configs.get(self.default_config, {})
+            else:
-        
+                result.update(raw)
-        # No thresholds configured
+        return result
-        return {}
     
     def check_value(
         self,
@@ -760,6 +931,12 @@ class ThresholdChecker:
         elif new_level == AlertLevel.WARNING and threshold.warning is not None:
             threshold_value = threshold.warning
 
+        # Keep hysteresis on the state so the UI can show the recovery threshold
+        if new_level != AlertLevel.OK:
+            alert_state.hysteresis = threshold.hysteresis
+        else:
+            alert_state.hysteresis = None
+
         # Update state and check for changes
         old_level = alert_state.level
         if alert_state.update(new_level, value, threshold_value, threshold.operator.value):
@@ -769,6 +946,36 @@ class ThresholdChecker:
             self._check_pending_or_renotify(host_name, alert_state, metric_path, value, threshold, None)
 
         return None
+    def _find_threshold(
+        self, thresholds: Dict[str, "ThresholdConfig"], metric_path: str
+    ) -> Tuple[Optional["ThresholdConfig"], Optional[str]]:
+        """Return (threshold, check_name) for *metric_path*, falling back to suffix matches.
+
+        Allows generic thresholds like ``nagios_runner.status_code`` to match
+        fully-qualified paths like ``nagios_runner.check_disk_root_status_code``.
+        The exact match is always tried first; then successive leading
+        underscore-delimited segments are stripped from the field name until
+        a match is found or no segments remain.
+
+        Returns:
+            (ThresholdConfig, None) for an exact match.
+            (ThresholdConfig, "check_disk_root") for a suffix match — the second
+            element is the stripped prefix, available as ``{check_name}`` in
+            display format templates.
+            (None, None) when no threshold is found.
+        """
+        if metric_path in thresholds:
+            return thresholds[metric_path], None
+        plugin, sep, field = metric_path.partition(".")
+        if not sep:
+            return None, None
+        parts = field.split("_")
+        for i in range(1, len(parts)):
+            candidate = plugin + "." + "_".join(parts[i:])
+            if candidate in thresholds:
+                return thresholds[candidate], "_".join(parts[:i])
+        return None, None
+
     def check_plugin_data(
         self,
         host_name: str,
@@ -797,11 +1004,10 @@ class ThresholdChecker:
         for metric_name, value in data.items():
             metric_path = f"{plugin_name}.{metric_name}"
 
-            if metric_path not in thresholds:
+            threshold, check_name = self._find_threshold(thresholds, metric_path)
+            if threshold is None:
                 continue
 
-            threshold = thresholds[metric_path]
-            
             # Get or create alert state
             if metric_path not in alert_states:
                 alert_states[metric_path] = AlertState(metric_path)
@@ -821,13 +1027,15 @@ class ThresholdChecker:
             elif new_level == AlertLevel.WARNING and threshold.warning is not None:
                 threshold_value = threshold.warning
 
+            alert_state.hysteresis = threshold.hysteresis if new_level != AlertLevel.OK else None
+
             # Update state and check for changes
             old_level = alert_state.level
             if alert_state.update(new_level, value, threshold_value, threshold.operator.value):
                 state_changes.append((metric_path, old_level, new_level, value))
-                self._apply_grace(host_name, alert_state, metric_path, old_level, new_level, value, threshold, data)
+                self._apply_grace(host_name, alert_state, metric_path, old_level, new_level, value, threshold, data, check_name=check_name, metric_name=metric_name)
             elif new_level != AlertLevel.OK:
-                self._check_pending_or_renotify(host_name, alert_state, metric_path, value, threshold, data)
+                self._check_pending_or_renotify(host_name, alert_state, metric_path, value, threshold, data, check_name=check_name, metric_name=metric_name)
 
         # Check nested metrics (e.g., partition data in disk_monitor)
         self._check_nested_metrics(
@@ -852,6 +1060,44 @@ class ThresholdChecker:
         # Get host-specific thresholds
         thresholds = self.get_thresholds_for_host(host_name)
         
+        # ZFS pool health checks
+        if plugin_name == "zfs_monitor" and "pools" in data:
+            pools = data["pools"]
+            if isinstance(pools, dict):
+                for pool_name, pool_metrics in pools.items():
+                    if not isinstance(pool_metrics, dict):
+                        continue
+                    # Synthesize status from health string for older clients
+                    # that predate the status field.
+                    pool_metrics_effective = dict(pool_metrics)
+                    if "health" in pool_metrics and "status" not in pool_metrics:
+                        pool_metrics_effective["status"] = 0 if pool_metrics["health"] == "ONLINE" else 1
+                    for metric_name, value in pool_metrics_effective.items():
+                        # Try specific pool name first, then wildcard '*'
+                        metric_path = f"{plugin_name}.{pool_name}.{metric_name}"
+                        wildcard_path = f"{plugin_name}.*.{metric_name}"
+                        threshold = thresholds.get(metric_path) or thresholds.get(wildcard_path)
+                        if threshold is None:
+                            continue
+                        if metric_path not in alert_states:
+                            alert_states[metric_path] = AlertState(metric_path)
+                        alert_state = alert_states[metric_path]
+                        new_level = threshold.evaluate_with_hysteresis(value, alert_state.level)
+                        threshold_value = None
+                        if new_level == AlertLevel.CRITICAL and threshold.critical is not None:
+                            threshold_value = threshold.critical
+                        elif new_level == AlertLevel.WARNING and threshold.warning is not None:
+                            threshold_value = threshold.warning
+                        alert_state.hysteresis = threshold.hysteresis if new_level != AlertLevel.OK else None
+                        pool_context = dict(pool_metrics_effective)
+                        pool_context["pool_name"] = pool_name
+                        old_level = alert_state.level
+                        if alert_state.update(new_level, value, threshold_value, threshold.operator.value):
+                            state_changes.append((metric_path, old_level, new_level, value))
+                            self._apply_grace(host_name, alert_state, metric_path, old_level, new_level, value, threshold, pool_context, metric_name=pool_name)
+                        elif new_level != AlertLevel.OK:
+                            self._check_pending_or_renotify(host_name, alert_state, metric_path, value, threshold, pool_context, metric_name=pool_name)
+
         # Look for partition data in disk_monitor
         if plugin_name == "disk_monitor" and "partitions" in data:
             partitions = data["partitions"]
@@ -887,6 +1133,8 @@ class ThresholdChecker:
                     elif new_level == AlertLevel.WARNING and threshold.warning is not None:
                         threshold_value = threshold.warning
 
+                    alert_state.hysteresis = threshold.hysteresis if new_level != AlertLevel.OK else None
+
                     old_level = alert_state.level
                     if alert_state.update(new_level, value, threshold_value, threshold.operator.value):
                         state_changes.append((metric_path, old_level, new_level, value))
@@ -903,6 +1151,8 @@ class ThresholdChecker:
         value: Any,
         threshold: ThresholdConfig,
         plugin_data: Optional[Dict[str, Any]] = None,
+        check_name: Optional[str] = None,
+        metric_name: Optional[str] = None,
     ):
         """Trigger a notification for an alert state change.
         
@@ -925,55 +1175,53 @@ class ThresholdChecker:
         # Format operator symbol
         op_symbol = threshold.operator.value
 
+        # Short metric label: strip the plugin-name prefix and _status_code suffix
+        short_path = (metric_path.partition(".")[2] or metric_path).removesuffix("_status_code")
+
         # Use a display-friendly value (inf is the sentinel for "overdue")
         import math
         display_value = "overdue" if isinstance(value, float) and math.isinf(value) else value
 
-        # Format message
+        # Format message — for the nagios operator there is no numeric threshold_value;
-        if new_level == AlertLevel.OK:
+        # render the display template whenever one is available.
-            lvl = "RECOVER"
+        has_display = threshold_value is not None or threshold.operator == ComparisonOperator.NAGIOS
-            message = f"{metric_path} = {display_value} ({old_level.name} -> OK)"
-        elif new_level == AlertLevel.WARNING:
-            lvl = "WARNING"
-            if threshold_value is not None:
-                threshold_info = self._format_display(
-                    threshold.display,
-                    value=display_value,
-                    threshold_value=threshold_value,
-                    op_symbol=op_symbol,
-                    plugin_data=plugin_data
-                )
-                message = f"{metric_path} = {display_value} {threshold_info}"
-            else:
-                message = f"{metric_path} = {display_value}"
-        elif new_level == AlertLevel.CRITICAL:
-            lvl = "CRITICAL"
-            if threshold_value is not None:
-                threshold_info = self._format_display(
-                    threshold.display,
-                    value=display_value,
-                    threshold_value=threshold_value,
-                    op_symbol=op_symbol,
-                    plugin_data=plugin_data
-                )
-                message = f"{metric_path} = {display_value} {threshold_info}"
-            else:
-                message = f"{metric_path} = {display_value}"
-        else:
-            lvl = "UNKNOWN"
-            message = f"{metric_path} = {display_value}"
 
-        # Return the formatted threshold info for storing in AlertState
+        def _fmt():
-        formatted_threshold_msg = None
+            return self._format_display(
-        if threshold_value is not None and new_level != AlertLevel.OK:
-            formatted_threshold_msg = self._format_display(
                 threshold.display,
                 value=display_value,
                 threshold_value=threshold_value,
                 op_symbol=op_symbol,
-                plugin_data=plugin_data
+                plugin_data=plugin_data,
+                check_name=check_name,
+                metric_name=metric_name,
             )
 
+        if new_level == AlertLevel.OK:
+            lvl = "RECOVER"
+            message = f"{short_path} = {display_value} ({old_level.name} -> OK)"
+        elif new_level == AlertLevel.WARNING:
+            lvl = "WARNING"
+            if has_display:
+                message = f"{short_path} = {display_value} {_fmt()}"
+            else:
+                message = f"{short_path} = {display_value}"
+        elif new_level == AlertLevel.CRITICAL:
+            lvl = "CRITICAL"
+            if has_display:
+                message = f"{short_path} = {display_value} {_fmt()}"
+            else:
+                message = f"{short_path} = {display_value}"
+        else:
+            lvl = "UNKNOWN"
+            if has_display:
+                message = f"{short_path} = {display_value} {_fmt()}"
+            else:
+                message = f"{short_path} = {display_value}"
+
+        # Formatted threshold info stored on AlertState for the UI
+        formatted_threshold_msg = _fmt() if has_display and new_level != AlertLevel.OK else None
+
         return lvl, message, formatted_threshold_msg
     
     def _send_notification(
@@ -987,19 +1235,29 @@ class ThresholdChecker:
         value: Any,
     ):
         """Send notification and log to journal/eventlog."""
+        from . import hbdclass
+        host = hbdclass.Host.hosts.get(host_name)
+        if host is not None and not host.watched:
+            eventlog(host_name, lvl, message, service="threshold")
+            return
+        short_path = (metric_path.partition(".")[2] or metric_path).removesuffix("_status_code")
+        title = f"[{lvl}] {host_name}  {short_path}"
+        # Strip the "metric = " prefix from message so body is just the value/detail
+        prefix = short_path + " = "
+        body = message[len(prefix):] if message.startswith(prefix) else message
         asyncio.get_event_loop().create_task(notify_mod.send_notification(
             host_name,
             notify_mod.Notification(
-                title=f"[{lvl}] {host_name}",
+                title=title,
-                body=message,
+                body=body,
                 level=lvl,
+                service=short_path,
             ),
         ))
         
         # Log to journal
         if self.journal is not None:
             try:
-                import asyncio
                 loop = asyncio.get_event_loop()
                 loop.create_task(self.journal.log_threshold_event(
                     host_name=host_name,
@@ -1017,33 +1275,62 @@ class ThresholdChecker:
         self,
         display_format: str,
         value: Any,
-        threshold_value: float,
+        threshold_value: Optional[float],
         op_symbol: str,
         plugin_data: Optional[Dict[str, Any]] = None,
+        check_name: Optional[str] = None,
+        metric_name: Optional[str] = None,
     ) -> str:
         """Format the display string using available data.
 
-        Args:
+        Available template variables:
-            display_format: Format string from threshold config
+            {value}           - current metric value
-            value: Current metric value
+            {threshold_value} - threshold that was exceeded
-            threshold_value: Threshold value that was exceeded
+            {op_symbol}       - comparison operator (>, <, >=, <=, ==, !=)
-            op_symbol: Comparison operator symbol
+            {check_name}      - prefix stripped for generic threshold match
-            plugin_data: Optional dictionary of plugin data fields
+                                (e.g. "check_disk_root" when metric
+                                "check_disk_root_status_code" matched generic
+                                threshold "status_code")
+            {metric_name}     - field name within the plugin data dict
+            Any key from plugin_data is also available.
 
         Returns:
             Formatted display string
         """
+        if not display_format:
+            display_format = "(threshold: {op_symbol} {threshold_value})" if threshold_value is not None else ""
+
         # Build format context with standard variables
         format_context = {
             'value': value,
-            'threshold_value': threshold_value,
             'op_symbol': op_symbol,
         }
+        if threshold_value is not None:
+            format_context['threshold_value'] = threshold_value
+
+        # Add generic-match context variables when available
+        if check_name is not None:
+            format_context['check_name'] = check_name
+        if metric_name is not None:
+            format_context['metric_name'] = metric_name
 
         # Add all plugin data fields if available
         if plugin_data:
             format_context.update(plugin_data)
 
+        # For nagios_runner generic matches, expose the matched check's output
+        # and status as short aliases {output} and {status} so display templates
+        # don't need to use the full {check_disk_root_output} form.
+        if check_name and plugin_data:
+            if 'output' not in format_context:
+                output = plugin_data.get(f"{check_name}_output")
+                if output is not None:
+                    format_context['output'] = output
+            if 'status' not in format_context:
+                status = plugin_data.get(f"{check_name}_status")
+                if status is not None:
+                    format_context['status'] = status
+        
         try:
             # Format the display string
             return display_format.format(**format_context)
@@ -1073,17 +1360,26 @@ class ThresholdChecker:
         value: Any,
         threshold: ThresholdConfig,
         plugin_data: Optional[Dict[str, Any]],
+        check_name: Optional[str] = None,
+        metric_name: Optional[str] = None,
     ) -> None:
         """Handle a state-change transition with grace-period logic.
 
-        Transitioning INTO alert: defers the notification for grace_seconds.
+        Transitioning INTO alert (worsening): defers the notification for the effective
+        grace period (threshold.grace if set, else self.grace_seconds). Grace of 0 fires
+        the notification immediately with no deferral.
+        De-escalation within alert states (e.g. CRITICAL→WARNING): no new notification;
+          the metric is still alerting so no RECOVER was sent.
         Transitioning TO OK:
           - Still in grace window (pending_since set): suppresses both the alert
             and the recovery — the spike never warranted a page.
           - Past grace: fires the RECOVER notification normally.
         """
+        effective_grace = threshold.grace if threshold.grace is not None else self.grace_seconds
+
         lvl, message, formatted_msg = self._trigger_notification(
-            host_name, metric_path, old_level, new_level, value, threshold, plugin_data
+            host_name, metric_path, old_level, new_level, value, threshold, plugin_data,
+            check_name=check_name, metric_name=metric_name,
         )
         alert_state.formatted_message = formatted_msg
 
@@ -1091,16 +1387,31 @@ class ThresholdChecker:
             if alert_state.pending_since is not None:
                 logger.info(
                     "Alert suppressed (recovered within %.0fs grace): %s on %s",
-                    self.grace_seconds, metric_path, host_name,
+                    effective_grace, metric_path, host_name,
                 )
                 alert_state.pending_since = None
             else:
                 self._send_notification(host_name, lvl, message, metric_path, old_level, new_level, value)
+        elif new_level.value > old_level.value:
+            # Worsening (OK→WARNING, OK→CRITICAL, WARNING→CRITICAL).
+            if effective_grace <= 0:
+                # No grace period — fire immediately.
+                self._send_notification(host_name, lvl, message, metric_path, old_level, new_level, value)
+                now = time.time()
+                alert_state.last_notification = now
+                alert_state.notification_count = 1
+            else:
+                alert_state.pending_since = time.time()
+                logger.debug(
+                    "Alert deferred (%.0fs grace): %s on %s = %s",
+                    effective_grace, metric_path, host_name, value,
+                )
         else:
-            alert_state.pending_since = time.time()
+            # De-escalation within alert states (e.g. CRITICAL→WARNING): metric is still
+            # alerting but did not recover, so no new notification.
             logger.debug(
-                "Alert deferred (%.0fs grace): %s on %s = %s",
+                "De-escalation %s→%s for %s on %s, no notification",
-                self.grace_seconds, metric_path, host_name, value,
+                old_level.name, new_level.name, metric_path, host_name,
             )
 
     def _check_pending_or_renotify(
@@ -1111,25 +1422,43 @@ class ThresholdChecker:
         value: Any,
         threshold: ThresholdConfig,
         plugin_data: Optional[Dict[str, Any]],
+        check_name: Optional[str] = None,
+        metric_name: Optional[str] = None,
     ) -> None:
         """Called when alert level is unchanged and non-OK.
 
         If a deferred notification is pending and grace_seconds have elapsed,
         fires it now. Otherwise falls through to normal reminder logic.
         """
+        effective_grace = threshold.grace if threshold.grace is not None else self.grace_seconds
         if alert_state.pending_since is not None:
-            if time.time() - alert_state.pending_since >= self.grace_seconds:
+            if time.time() - alert_state.pending_since >= effective_grace:
                 lvl, message, formatted_msg = self._trigger_notification(
-                    host_name, metric_path, AlertLevel.OK, alert_state.level, value, threshold, plugin_data
+                    host_name, metric_path, AlertLevel.OK, alert_state.level, value, threshold, plugin_data,
+                    check_name=check_name, metric_name=metric_name,
                 )
                 alert_state.formatted_message = formatted_msg
                 self._send_notification(
                     host_name, lvl, message, metric_path, AlertLevel.OK, alert_state.level, value
                 )
                 alert_state.pending_since = None
+                now = time.time()
+                alert_state.last_notification = now
+                alert_state.notification_count = 1
             # else: still within grace window, do nothing
         else:
-            self._check_renotify(host_name, alert_state, metric_path, value, threshold, plugin_data)
+            self._check_renotify(host_name, alert_state, metric_path, value, threshold, plugin_data, check_name=check_name, metric_name=metric_name)
+
+    @staticmethod
+    def _human_duration(seconds: float) -> str:
+        s = int(seconds)
+        if s < 120:
+            return f"{s}s"
+        if s < 3600:
+            return f"{s // 60}m {s % 60}s"
+        h, rem = divmod(s, 3600)
+        m = rem // 60
+        return f"{h}h {m}m" if m else f"{h}h"
 
     def _check_renotify(
         self,
@@ -1139,6 +1468,8 @@ class ThresholdChecker:
         value: Any,
         threshold: ThresholdConfig,
         plugin_data: Optional[Dict[str, Any]] = None,
+        check_name: Optional[str] = None,
+        metric_name: Optional[str] = None,
     ):
         """Check if we should send a repeat notification.
         
@@ -1176,6 +1507,7 @@ class ThresholdChecker:
             
             # Format operator symbol
             op_symbol = threshold.operator.value
+            short_path = (metric_path.partition(".")[2] or metric_path).removesuffix("_status_code")
 
             # Time to re-notify
             if threshold_value is not None:
@@ -1185,23 +1517,63 @@ class ThresholdChecker:
                     value=value,
                     threshold_value=threshold_value,
                     op_symbol=op_symbol,
-                    plugin_data=plugin_data
+                    plugin_data=plugin_data,
+                    check_name=check_name,
+                    metric_name=metric_name,
                 )
-                message = f"REMINDER ({alert_state.level.name}): {host_name} - {metric_path} = {value} {threshold_info}, ongoing for {int(now - alert_state.since)}s"
+                body = f"{value} {threshold_info}, ongoing for {self._human_duration(now - alert_state.since)}"
             else:
-                message = f"REMINDER ({alert_state.level.name}): {host_name} - {metric_path} = {value} (ongoing for {int(now - alert_state.since)}s)"
+                body = f"{value} (ongoing for {self._human_duration(now - alert_state.since)})"
+            message = f"REMINDER ({alert_state.level.name}): {host_name} - {short_path} = {body}"
 
-            asyncio.get_event_loop().create_task(notify_mod.send_notification(
+            from . import hbdclass
-                host_name,
+            host = hbdclass.Host.hosts.get(host_name)
-                notify_mod.Notification(
+            if host is None or host.watched:
-                    title=f"[REMINDER/{alert_state.level.name}] {host_name}",
+                asyncio.get_event_loop().create_task(notify_mod.send_notification(
-                    body=message,
+                    host_name,
-                    level=alert_state.level.name,
+                    notify_mod.Notification(
-                ),
+                        title=f"[REMINDER/{alert_state.level.name}] {host_name}  {short_path}",
-            ))
+                        body=body,
+                        level=alert_state.level.name,
+                        service=short_path,
+                    ),
+                ))
+                logger.info("Re-notification sent: %s", message)
             alert_state.last_notification = now
             alert_state.notification_count += 1
-            logger.info("Re-notification sent: %s", message)
+    
+    def purge_stale_alerts(self, hbdclass) -> None:
+        """Remove alert states that have no matching threshold configuration.
+
+        Called after startup (pickle restore) and after each config reload so
+        that alerts orphaned by configuration changes do not linger forever.
+        Alerts whose metric_path is not present in the current threshold config
+        for that host are silently dropped.
+        """
+        for hostname, host in hbdclass.Host.hosts.items():
+            if not host.alert_states:
+                continue
+            configured = self.get_thresholds_for_host(hostname)
+            stale = []
+            for mp in host.alert_states:
+                # connectivity.* and rtt are managed by the connection state
+                # machine, not by threshold config — never purge them.
+                if mp == "rtt" or mp.startswith("connectivity"):
+                    continue
+                if self._find_threshold(configured, mp)[0] is not None:
+                    continue
+                # Also match wildcard pool/partition thresholds (e.g. "zfs_monitor.*.status"
+                # covers alert state "zfs_monitor.tank.status").
+                parts = mp.split(".")
+                if len(parts) == 3 and f"{parts[0]}.*.{parts[2]}" in configured:
+                    continue
+                stale.append(mp)
+            for mp in stale:
+                logger.info(
+                    "Purging stale alert state for %s / %s (no threshold configured)",
+                    hostname, mp,
+                )
+                del host.alert_states[mp]
 
     def get_active_alerts(self, alert_states: Dict[str, AlertState]) -> list:
         """

hbd/server/udp.py

@@ -16,6 +16,11 @@ from . import notify as notify_mod
 logger = logging.getLogger(__name__)
 eventlog = notify_mod.eventlog
 
+# Plugin data is declared stale after this many collection intervals without a
+# fresh sample. >= 2 so a single missed sample never purges live data (the user
+# directive: wait at least two full intervals after recovery before going stale).
+_STALE_INTERVAL_MULTIPLIER = 3
+
 # SO_TIMESTAMP: kernel attaches a struct timeval to each received datagram.
 # Supported on Linux, FreeBSD, and macOS.  The constant is not exposed by
 # Python's socket module on all platforms 
@@ -211,10 +216,11 @@ def _make_timer_callbacks(uname, host, ctx):
         connection.newstate(connection.__class__.OVERDUE, now, cfg.get("grace", 2))
         msg = f"{connection.afam} overdue"
         eventlog(uname, "CRITICAL", msg)
-        asyncio.create_task(notify_mod.send_notification(
+        if host.watched:
-            uname,
+            asyncio.create_task(notify_mod.send_notification(
-            notify_mod.Notification(title=f"[CRITICAL] {uname}", body=msg, level="CRITICAL"),
+                uname,
-        ))
+                notify_mod.Notification(title=f"[CRITICAL] {uname}", body=msg, level="CRITICAL"),
+            ))
         # Track in alert_states so the Alerts Dashboard shows this
         _set_connectivity_alert(host, connection.afam, "CRITICAL")
         if threshold_checker:
@@ -231,6 +237,23 @@ def _make_timer_callbacks(uname, host, ctx):
     return on_overdue, on_unknown
 
 
+def _make_plugin_stale_callback(uname, ctx):
+    """Return an async callback that clears stale plugin data and its alerts."""
+    msg_to_websockets = ctx.get("msg_to_websockets")
+
+    async def on_plugin_stale(host, plugin_name):
+        host.plugin_data.pop(plugin_name, None)
+        stale_keys = [k for k in host.alert_states if k.startswith(f"{plugin_name}.")]
+        for k in stale_keys:
+            del host.alert_states[k]
+        eventlog(uname, "INFO", f"plugin data stale: {plugin_name}")
+        if msg_to_websockets:
+            msg_to_websockets("plugin_stale", {"host": uname, "plugin": plugin_name})
+            msg_to_websockets("host", host.stateinfo())
+
+    return on_plugin_stale
+
+
 def restore_connection_timers(hbdclass, ctx):
     """Restore overdue timers for all loaded connections after a pickle restore.
 
@@ -248,10 +271,15 @@ def restore_connection_timers(hbdclass, ctx):
         for afam, conn in list(host.connections.items()):
             state = conn.getstate()
             if state == hbdclass.Connection.DOWN:
+                _set_connectivity_alert(host, afam, "CRITICAL")
                 continue
 
             on_overdue, on_unknown = _make_timer_callbacks(uname, host, ctx)
 
+            if state == hbdclass.Connection.UNKNOWN:
+                _set_connectivity_alert(host, afam, "CRITICAL")
+                continue
+
             if state == hbdclass.Connection.UP and interval > 0:
                 elapsed = now - conn.lastbeat
                 # Give hosts one full (interval + grace) of extra time on startup
@@ -282,6 +310,10 @@ def restore_connection_timers(hbdclass, ctx):
                         "Restored OVERDUE timer %s/%s: %.0fs remaining",
                         uname, afam, remaining,
                     )
+                # Ensure the connectivity alert is set — it may be missing if
+                # hbd was shut down before the on_overdue callback had a chance
+                # to record it.
+                _set_connectivity_alert(host, afam, "CRITICAL")
                 restored += 1
 
     logger.info("Restored timers for %d connection(s)", restored)
@@ -332,11 +364,12 @@ def handle_datagram(msg: dict, addr, transport, ctx: dict):
         # Use new config function to check dyndns
         dyndnshosts = config_mod.get_dyndnshosts(cfg)
         host.dyn = uname in dyndnshosts
+        watchhosts = config_mod.get_watchhosts(cfg)
+        host.watched = uname in watchhosts
         # Apply user-access settings from config
         access = config_mod.get_host_access(cfg, uname)
         host.apply_access(access["owner"], access["managers"], access["monitors"])
-        if verbose:
+        logger.info("New host signed on: %s (dyn=%s, access=%s)", uname, host.dyn, access)
-            print(("XX: New host, num now %s" % (len(hbdcls.Host.hosts))))
         newh = True
     else:
         host = hbdcls.Host.hosts[uname]
@@ -350,8 +383,10 @@ def handle_datagram(msg: dict, addr, transport, ctx: dict):
 
     if msg.get("ID") == "HTB":
         host.doesack = msg.get("acks", -1)
-        # send ACK back
+        # send ACK back; ask client to resend plugin info when we have none yet
         rmsg = {"time": time.time()}
+        if not host.plugin_data:
+            rmsg["request_update"] = 1
         opkt = dicttos("ACK", rmsg)
         try:
             transport.sendto(opkt, addr)
@@ -364,10 +399,52 @@ def handle_datagram(msg: dict, addr, transport, ctx: dict):
         plugin_name = msg.get("plugin")
         if plugin_name:
             # Extract plugin fields, dropping protocol metadata fields
+            # (_interval is the client-declared collection interval, not a metric).
             plugin_data = {k: v for k, v in msg.items()
-                           if k not in ("ID", "plugin", "id", "name")}
+                           if k not in ("ID", "plugin", "id", "name", "_interval")}
             # Store plugin data with timestamp
             host.add_plugin_data(plugin_name, plugin_data, timestamp=now)
+
+            # Set the stale timer. Prefer the interval the client declares: it is
+            # the true collection cadence and is correct from the very first
+            # post-recovery sample, so we never purge data too early after an
+            # outage. interval == 0 means a collect-once InfoPlugin that never
+            # goes stale. Legacy clients omit _interval, so fall back to inferring
+            # the cadence from the gap between the last two received samples.
+            history = host.plugin_data.get(plugin_name, [])
+            declared = msg.get("_interval")
+            if declared is not None:
+                if declared > 0:
+                    host.reset_plugin_timer(plugin_name, declared * _STALE_INTERVAL_MULTIPLIER,
+                                            _make_plugin_stale_callback(uname, ctx))
+                else:
+                    host.cancel_plugin_timer(plugin_name)
+            elif len(history) >= 2:
+                plugin_interval = max(history[-1][0] - history[-2][0], 1)
+                host.reset_plugin_timer(plugin_name, plugin_interval * _STALE_INTERVAL_MULTIPLIER,
+                                        _make_plugin_stale_callback(uname, ctx))
+            else:
+                host.cancel_plugin_timer(plugin_name)
+
+            # Remove alert states for metrics present in the previous sample but
+            # absent now (e.g. a nagios check removed from configuration).
+            if len(history) >= 2:
+                prev_keys = set(history[-2][1].keys())
+                curr_keys = set(plugin_data.keys())
+                for metric_name in prev_keys - curr_keys:
+                    metric_path = f"{plugin_name}.{metric_name}"
+                    if host.alert_states.pop(metric_path, None) is not None:
+                        eventlog(uname, "INFO", f"stale check removed: {metric_path}")
+                if (prev_keys - curr_keys) and msg_to_websockets:
+                    msg_to_websockets("host", host.stateinfo())
+
+            # If os_info reports an owner and none is configured server-side, apply it
+            if plugin_name == "os_info":
+                config_owner =  config_mod.get_host_access(cfg, uname).get("owner")
+                default_owner = config_mod.get_default_owner(cfg)
+                inferred_owner = config_owner or plugin_data.get("owner") or default_owner
+                host.owner = inferred_owner
+                logger.info(f"owner for {uname} is {host.owner}")
             if DEBUG > 1:
                 print(f"Stored plugin data for {uname}: {plugin_name}")
             
@@ -407,10 +484,11 @@ def handle_datagram(msg: dict, addr, transport, ctx: dict):
 
     if res:
         eventlog(uname, "WARNING", res)
-        asyncio.create_task(notify_mod.send_notification(
+        if host.watched:
-            uname,
+            asyncio.create_task(notify_mod.send_notification(
-            notify_mod.Notification(title=f"[WARNING] {uname}", body=res, level="WARNING"),
+                uname,
-        ))
+                notify_mod.Notification(title=f"[WARNING] {uname}", body=res, level="WARNING"),
+            ))
 
     interval = int(msg.get("interval", 0) or 0)
     shutdown = msg.get("shutdown", 0)
@@ -419,17 +497,32 @@ def handle_datagram(msg: dict, addr, transport, ctx: dict):
     boot = msg.get("boot", 0)
 
     if boot:
+        # hbc was stared with a -b flag
         eventlog(uname, "INFO", "booted")
-        asyncio.create_task(notify_mod.send_notification(
+        if host.watched:
-            uname,
+            asyncio.create_task(notify_mod.send_notification(
-            notify_mod.Notification(title=f"[INFO] {uname}", body=f"{host.name} booted", level="INFO"),
+                uname,
-        ))
+                notify_mod.Notification(title=f"[INFO] {uname}", body=f"{host.name} booted", level="INFO"),
+            ))
     if message:
-        eventlog(uname, "INFO", "msg: %s" % message, service=service)
+        eventlog(uname, "INFO", message, service=service)
 
     if conn.getstate() != hbdcls.Connection.UP:
+        # Transition to UP and log/notify if appropriate
         lasts = conn.state
         d = conn.newstate(hbdcls.Connection.UP, now)
+        # On reboot, pre-boot plugin data and derived alerts are stale.
+        # Cancel all plugin timers and wipe plugin state so timers restart
+        # cleanly from the first two post-boot samples.
+        for pname in list(host.plugin_timers):
+            host.cancel_plugin_timer(pname)
+        host.plugin_data.clear()
+        stale_plugin_keys = [
+            k for k in host.alert_states
+            if k not in ("rtt",) and not k.startswith("connectivity.")
+        ]
+        for k in stale_plugin_keys:
+            del host.alert_states[k]
         # Clear connectivity alert now that the host is back up
         _set_connectivity_alert(host, conn.afam, "OK")
         # Don't log/notify RECOVER for a brand-new host seen for the first time —
@@ -437,13 +530,18 @@ def handle_datagram(msg: dict, addr, transport, ctx: dict):
         if not newh:
             if d == 0 or lasts == "unknown":
                 m = "%s is up" % (conn.afam)
+            elif d < 4:
+                # Transient blip (likely client restart) — skip log and notification
+                m = None
             else:
                 m = "%s back after being %s for %s" % (conn.afam, lasts, dur(d))
-            eventlog(uname, "RECOVER", m)
+            if m:
-            asyncio.create_task(notify_mod.send_notification(
+                eventlog(uname, "RECOVER", m)
-                uname,
+                if host.watched:
-                notify_mod.Notification(title=f"[RECOVER] {uname}", body=m, level="RECOVER"),
+                    asyncio.create_task(notify_mod.send_notification(
-            ))
+                        uname,
+                        notify_mod.Notification(title=f"[RECOVER] {uname}", body=m, level="RECOVER"),
+                    ))
 
     if boot or newh:
         host.upcount = host.doesack
@@ -453,10 +551,11 @@ def handle_datagram(msg: dict, addr, transport, ctx: dict):
     if shutdown:
         m = "%s shutdown" % conn.afam
         eventlog(uname, "INFO", m)
-        asyncio.create_task(notify_mod.send_notification(
+        if host.watched:
-            uname,
+            asyncio.create_task(notify_mod.send_notification(
-            notify_mod.Notification(title=f"[INFO] {uname}", body=m, level="INFO"),
+                uname,
-        ))
+                notify_mod.Notification(title=f"[INFO] {uname}", body=m, level="INFO"),
+            ))
         conn.newstate(hbdcls.Connection.DOWN, now)
         _set_connectivity_alert(host, conn.afam, "CRITICAL")
 

hbd/server/users.py

@@ -146,9 +146,14 @@ def load_users(config: dict) -> dict:
     Returns the new ``users`` dict.
     """
     global users
+    old_users = dict(users)  # snapshot before rebuild
     users_cfg = config.get("users", {})
     if not isinstance(users_cfg, dict):
         users = {}
+        # Preserve OAuth-provisioned users (password_hash == "") that aren't in config.
+        for username, existing_user in old_users.items():
+            if not existing_user.password_hash and username not in users:
+                users[username] = existing_user
         return users
 
     result: dict = {}
@@ -166,6 +171,10 @@ def load_users(config: dict) -> dict:
         )
 
     users = result
+    # Preserve OAuth-provisioned users (password_hash == "") that aren't in config.
+    for username, existing_user in old_users.items():
+        if not existing_user.password_hash and username not in users:
+            users[username] = existing_user
     logger.info("Loaded %d user(s) from config", len(users))
     return users
 
@@ -187,6 +196,26 @@ def authenticate(username: str, password: str) -> "User | None":
     return None
 
 
+def provision_oauth_user(username: str, full_name: str, avatar: str) -> "User":
+    """Create or update a user sourced from an OAuth2 provider.
+
+    New users are inserted with no password_hash — they can only authenticate
+    via OAuth.  Existing users (e.g. defined in config with a password) have
+    their display name and avatar refreshed; all other attributes are preserved.
+    """
+    user = users.get(username)
+    if user is None:
+        user = User(username=username, full_name=full_name, avatar=avatar)
+        users[username] = user
+        logger.info("Provisioned OAuth user %r", username)
+    else:
+        if full_name:
+            user.full_name = full_name
+        if avatar:
+            user.avatar = avatar
+    return user
+
+
 # ---------------------------------------------------------------------------
 # Session management
 # ---------------------------------------------------------------------------

hbd/server/ws.py

@@ -13,7 +13,8 @@ from . import data
 
 logger = logging.getLogger(__name__)
 
-_connections: set = set()
+# Map of WebSocket → User object (or None when auth is disabled)
+_connections: dict = {}
 _loop: Optional[asyncio.AbstractEventLoop] = None
 _get_hosts: Optional[Callable[[], Iterable]] = None
 _verbose: bool = False
@@ -34,31 +35,65 @@ def setup(
     _verbose = verbose
 
 
+def _user_can_see_host(user, host_name: str) -> bool:
+    """Return True if *user* may see updates for *host_name* (manager or higher)."""
+    from . import hbdclass, users as users_mod
+    if user is None or not users_mod.users_enabled():
+        return True
+    if user.admin:
+        return True
+    host = hbdclass.Host.hosts.get(host_name)
+    if host is None:
+        return False
+    return host.is_manager(user.username)
+
+
+def _get_token(request) -> str:
+    """Extract session token from request (mirrors logic in http.py)."""
+    auth = request.headers.get("Authorization", "")
+    if auth.startswith("Bearer "):
+        return auth[7:].strip()
+    token = request.headers.get("X-Auth-Token", "")
+    if token:
+        return token
+    return request.cookies.get("hbd_session", "")
+
+
 async def handler(request):
     """aiohttp WebSocket upgrade handler — register as GET /ws."""
     from aiohttp import web
+    from . import users as users_mod
 
     ws = web.WebSocketResponse()
     await ws.prepare(request)
 
-    _connections.add(ws)
+    token = _get_token(request)
+    user = users_mod.get_session_user(token) if token else None
+
+    _connections[ws] = user
     remote = request.remote
     logger.info("WebSocket connected from %s", remote)
 
     try:
-        # Send current host state to the new client
+        # Send current host state, filtered to hosts this user may see
         if _get_hosts:
             try:
                 for h in list(_get_hosts()):
-                    await ws.send_str(json.dumps({"type": "host", "data": h}))
+                    host_name = h.get("raw_name") or h.get("name", "")
+                    if _user_can_see_host(user, host_name):
+                        await ws.send_str(json.dumps({"type": "host", "data": h}))
             except Exception as e:
                 logger.error("Error sending initial hosts: %s", e)
 
-        # Send recent messages
+        # Send recent messages newest-first so the client can append them in
+        # display order without reordering on arrival (tagged history=True so
+        # the client knows to append rather than prepend).
         if data.msgs:
             try:
-                for m in data.msgs:
+                for m in reversed(data.msgs[-30:]):
-                    await ws.send_str(json.dumps({"type": "message", "data": m}))
+                    host_name = m.get("host") if isinstance(m, dict) else None
+                    if not host_name or _user_can_see_host(user, host_name):
+                        await ws.send_str(json.dumps({"type": "message", "data": m, "history": True}))
             except Exception as e:
                 logger.error("Error sending initial messages: %s", e)
 
@@ -74,7 +109,7 @@ async def handler(request):
     except Exception as e:
         logger.exception("WebSocket handler error from %s: %s", remote, e)
     finally:
-        _connections.discard(ws)
+        _connections.pop(ws, None)
         logger.info("WebSocket disconnected from %s", remote)
 
     return ws
@@ -83,25 +118,39 @@ async def handler(request):
 def broadcast(typ: str, payload) -> bool:
     """Thread-safe broadcast to all connected WebSocket clients.
 
+    For host and plugin updates, only sends to clients whose user has
+    manager-or-higher access to that host.  Other message types are
+    broadcast to all clients.
+
     Can be called from any thread; schedules sends on the event loop.
     Returns False if the loop is not running yet.
     """
     if not _loop:
         return False
+
+    # Determine the host name for access-filtered message types
+    host_name: Optional[str] = None
+    if typ in ("host", "plugin"):
+        host_name = payload.get("raw_name") or payload.get("host") or payload.get("name")
+    elif typ == "message" and isinstance(payload, dict):
+        host_name = payload.get("host")
+
     jmsg = json.dumps({"type": typ, "data": payload})
 
     async def _send_all():
         dead = set()
-        for ws in list(_connections):
+        for ws, user in list(_connections.items()):
             try:
-                if not ws.closed:
+                if ws.closed:
-                    await ws.send_str(jmsg)
-                else:
                     dead.add(ws)
+                    continue
+                if host_name is not None and not _user_can_see_host(user, host_name):
+                    continue
+                await ws.send_str(jmsg)
             except Exception:
                 dead.add(ws)
         for ws in dead:
-            _connections.discard(ws)
+            _connections.pop(ws, None)
 
     asyncio.run_coroutine_threadsafe(_send_all(), _loop)
     return True

pyproject.toml

@@ -4,20 +4,32 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "hbd"
-version = "5.1.9"
+version = "5.4.0"
 description = "Heartbeat monitoring system — client (hbc) and server (hbd)"
 readme = "README.md"
 requires-python = ">=3.11"
-license = "MIT" 
-keywords = ["heartbeat", "monitoring", "dns", "websocket", "system-monitoring"]
-authors = [
-  { name = "heartbeat contributors" }
-]
-
-# Core dependencies (required for both client and server)
 dependencies = [
   "PyYAML>=6.0",
 ]
+license = "MIT"
+license-files = ["LICENSE.md"]
+keywords = ["heartbeat", "monitoring", "dns", "websocket", "system-monitoring"]
+authors = [
+  { name = "Andreas Wrede" }
+]
+classifiers = [
+  "Programming Language :: Python :: 3",
+  "Programming Language :: Python :: 3.11",
+  "Programming Language :: Python :: 3.12",
+  "Programming Language :: Python :: 3.13",
+  "Operating System :: POSIX :: Linux",
+  "Operating System :: POSIX :: BSD",
+  "Topic :: System :: Monitoring",
+  "Topic :: System :: Networking :: Monitoring",
+]
+
+[project.urls]
+Repository = "https://git.wrede.ca/andreas/heartbeat"
 
 [project.optional-dependencies]
 # Client-only dependencies (hbc - system monitoring client)
@@ -32,6 +44,7 @@ server = [
   "aiohttp>=3.11",
   "Jinja2>=3.1.6",
   "matrix-nio>=0.24",
+  "ruamel.yaml>=0.18",
 ]
 
 # Minimal client — hbc_mini only, no external dependencies

rndc-key

@@ -1,4 +0,0 @@
-key "rndc-key" {
-      algorithm hmac-md5;
-      secret "qlGa+AYKtyOgWNuozqECMw==";
-};

scripts/bumpminor.sh

@@ -4,12 +4,29 @@ set -e
 uv version --bump patch 
 VER=$(uv  version  --short)
 sed -i".bak"  "s/__version__ = \"[0-9.]*\"\(.*\)$/__version__ = \"$VER\"\1/" hbd/__init__.py
+sed -i".bak"  "s/__version__ = \"[0-9.]*\"\(.*\)$/__version__ = \"$VER\"\1/" scripts/hbc_mini.py
+sed -i".bak"  "s/\*\*Package:\*\* \`hbd\` v[0-9.]*/\*\*Package:\*\* \`hbd\` v$VER/" README.md
+
+# Update CHANGELOG.md with commits since last tag
+LASTTAG=$(git describe --tags --abbrev=0 2>/dev/null || true)
+ADDED=$(git log "${LASTTAG:+$LASTTAG..}HEAD" --pretty="%s" | grep "^feat:" | sed 's/^feat: /- /')
+FIXED=$(git log "${LASTTAG:+$LASTTAG..}HEAD" --pretty="%s" | grep "^fix:" | sed 's/^fix: /- /')
+{
+    printf "## [%s]\n" "$VER"
+    [ -n "$ADDED" ] && printf "\n### Added\n%s\n" "$ADDED"
+    [ -n "$FIXED" ] && printf "\n### Fixed\n%s\n" "$FIXED"
+    printf "\n---\n\n"
+} > /tmp/changelog_entry.txt
+sed -i".bak" "4r /tmp/changelog_entry.txt" CHANGELOG.md
+rm /tmp/changelog_entry.txt CHANGELOG.md.bak
 
 # commit pyproject.toml
-git commit -m "version $VER" pyproject.toml hbd/__init__.py
+git commit -m "version $VER" pyproject.toml hbd/__init__.py scripts/hbc_mini.py README.md CHANGELOG.md
 git push 
 # tag version
 git tag -a v$VER -m "Version $VER"
 git push --tags
 
 rm hbd/__init__.py.bak
+rm scripts/hbc_mini.py.bak
+rm README.md.bak

scripts/c/.gitignore

@@ -0,0 +1,2 @@
+hbc_mini
+hbc_mini_dbg

scripts/c/Makefile

@@ -0,0 +1,21 @@
+CC      ?= cc
+CFLAGS  = -O2 -Wall -Wextra -std=c11
+LDFLAGS = -lz -lpthread -lm
+TARGET  = hbc_mini
+SRC     = hbc_mini.c
+
+# FreeBSD/NetBSD keep zlib in base; no extra flags needed.
+# On some NetBSD installs pthreads may need -lpthread from pkgsrc.
+
+.PHONY: all clean debug
+
+all: $(TARGET)
+
+$(TARGET): $(SRC)
+	$(CC) $(CFLAGS) -o $@ $< $(LDFLAGS)
+
+debug: $(SRC)
+	$(CC) -g -fsanitize=address,undefined -o $(TARGET)_dbg $< $(LDFLAGS)
+
+clean:
+	rm -f $(TARGET) $(TARGET)_dbg

scripts/hb_install.sh

@@ -17,9 +17,9 @@ venv=""
 [ "$2" = "HA" ] && on_ha=1
 [ -z "$what" ] && what="client"
 
-if [ -d /homeassistant ]; then
+if [ -d /homeassistant ]; then  # if running from HA command line
-    echo "HA, running \"docker exec homeassistant $0 $@\""
+    echo "HA, running \"docker exec homeassistant /config/bin/hb_install.sh $@\""
-    docker exec homeassistant $0 $@ HA
+    docker exec homeassistant /config/bin/hb_install.sh $@ HA
     rc=$?
     if [ $rc -ne 0 ]; then
         echo "Failed to install heartbeat in HA, please check the logs for more details"
@@ -28,8 +28,9 @@ if [ -d /homeassistant ]; then
     exit 0
 fi
 
-if [ $on_ha -eq 1 ]; then
+if [ $on_ha -eq 1 ] || [ -r /.dockerenv ] && [ -d /config/bin ]; then
-    echo "Installing under docker on Home Assistant OS, using /config/bin for executables and /config/venvs for virtual environments "
+    # Installing under docker on Home Assistant OS, using /config/bin for executables and /config/venvs for virtual environments 
+    echo "Home Assistant OS detected, installing under docker"
     where="/config/bin"
     venv="/config/venvs"
 else
@@ -52,23 +53,26 @@ else
         venv="$HOME/venvs"
     fi
 fi
-
+echo "Installing $what to $where"
-echo "Installing heartbeat $what"
+if [ ! -z "$venv" ]; then
+    echo "Using virtual environment at $venv/hbd"
+fi
 
 if [ "$venv" != "" ] && [ ! -d  $venv/hbd ]; then
-    set +e
-    python3 -m pip --version > /dev/null 2>&1 
-    rc=$?
-    set -e
     arg=""
-    if [ $rc -ne 0 ]; then
+    have_pip=$(python3 -c "import pip" 2>/dev/null &> /dev/null && echo "Installed" || echo "Not Installed")
+    if [ "$have_pip" = "Not Installed" ]; then
         # some systems do not have pip installed by default, so we need to fetch get-pip.py and install pip
         echo "pip is not installed, fetching get-pip.py and installing pip"
         arg="--without-pip"
     fi
     mkdir -p $venv
-    have_venv=$(python3 -c "import venv" &> /dev/null && echo "Installed" || echo "Not Installed")
+    have_venv=$(python3 -c "import venv" 2>/dev/null &> /dev/null && echo "Installed" || echo "Not Installed")
     if [ "$have_venv" = "Not Installed" ]; then
+        if [ "$have_pip" = "Not Installed" ]; then
+            echo "python has no venv, and no pip to install virtualenv, cannot continue"
+            exit 1
+        fi
         echo "python venv module not found, installing virtualenv"
         python3 -m pip install --user virtualenv
         python3 -m virtualenv $venv/hbd --system-site-packages $arg
@@ -82,43 +86,30 @@ if [ "$venv" != "" ] && [ ! -d  $venv/hbd ]; then
     deactivate
 fi
 
-if [ -z "$venv" ]; then
+if [ ! -z "$venv" ]; then
-    echo "Installing heartbeat $what globally"
-else
-    echo "Installing heartbeat $what in virtual environment $venv/hbd"
     . $venv/hbd/bin/activate
 fi
-python3 -mpip install --upgrade --index-url https://git.wrede.ca/api/packages/andreas/pypi/simple/ --extra-index-url https://pypi.org/simple hbd[$what]
+if [ "$what" = "mini" ]; then
+    curl -s -o $where/hbc_mini https://git.wrede.ca/andreas/heartbeat/raw/branch/master/scripts/hbc_mini.py
+    chmod +x $where/hbc_mini
+else
+    python3 -mpip install --upgrade --index-url https://git.wrede.ca/api/packages/andreas/pypi/simple/ --extra-index-url https://pypi.org/simple hbd[$what]
+fi
 
-if [ "$what" = "server" ]; then
+if [ ! -z "$venv" ]; then
-    rm -f $where/hbd
+    echo "linking executables to $where"
-    ln -sf $(which hbd) $where/hbd
+    if [ "$what" = "server" ]; then
-    echo "hbd installed, you can run it with \"$where/hbd\" or \"hbd\" if $where is in your PATH"
+        rm -f $where/hbd
-elif [ "$what" = "client" ]; then
+        ln -sf $(which hbd) $where/hbd
-    hbc_path=$(which hbc)
+    elif [ "$what" = "client" ]; then
-    if [ -z "$hbc_path" ]; then
-        echo "hbc not found in PATH, installation failed"
-        exit 1
-    fi
-    if [ "$hbc_path" != "$where/hbc" ]; then
         rm -f $where/hbc
         ln -sf $(which hbc) $where/hbc
     fi
-    if [ "$0" != "$where/hb_install.sh" ]; then
+    rm -f $where/hb_install.sh
-        cp "$0" $where/hb_install.sh
+    ln -sf $(which hb_install.sh) $where/hb_install.sh
-        chmod +x $where/hb_install.sh
-    fi
-    if [ $on_ha -eq 1 ]; then
-        echo "restarting hbc "
-        job=$(grep run_hbc configuration.yaml | sed 's/run_hbc://')
-        $job
-    else
-        echo "hbc installed, you can run it with \"$where/hbc\" or \"hbc\" if $where is in your PATH"
-    fi  
-elif [ "$what" = "mini" ]; then
-    hbc_path=$(which hbc_mini)
-    if [ "$hbc_path" != "$where/hbc_mini" ]; then
-       ln -sf $(which hbc) $where/hbc_mini
-    fi
-    echo "hbc mini installed, you can run it with \"$where/hbc_mini\" or \"hbc_mini\" if $where is in your PATH"
 fi
+echo "Installation complete. To upgrade, run the following:"
+echo "    $where/hb_install.sh $what"
+echo "To install on another machine, run the following obtain the install script and run it:"
+echo "from https://git.wrede.ca/andreas/heartbeat/raw/branch/master/scripts/hb_install.sh"
+echo "and then run sh hb_install.sh [mini|client]"

scripts/hbc_mini.py

@@ -40,6 +40,9 @@ from logging.handlers import SysLogHandler
 from pathlib import Path
 from typing import Any, Dict, List, Optional, Tuple
 
+# updated by scripts/bumpminor.sh
+__version__ = "5.4.0"
+
 # ---------------------------------------------------------------------------
 # Protocol  (mirrors hbd/common/proto.py)
 # ---------------------------------------------------------------------------
@@ -111,9 +114,15 @@ def _stodict(data: bytes) -> Dict[str, Any]:
 _DEFAULTS: Dict[str, Any] = {
     "hb_port": 50003,
     "interval": 10,
+    "owner": None,
+    "allow_remote_command": False,  # Execute shell commands received in CMD packets
     "plugins": {},
 }
 
+# Set from config in main(). Off by default: a CMD packet is an unauthenticated
+# UDP datagram, so executing one must be opted into per host.
+_allow_remote_command = False
+
 
 def _load_config(path: Optional[str] = None) -> Dict[str, Any]:
     cfg = dict(_DEFAULTS)
@@ -233,9 +242,11 @@ class OSInfoPlugin(InfoPlugin):
             "machine": platform.machine(),
             "architecture": platform.architecture()[0],
             "python_version": platform.python_version(),
-            "hbc_version": "5.1.8",
+            "hbc_version": __version__,
             "hbc_type": "mini",
         }
+        if self.config.get("owner"):
+            data["owner"] = self.config["owner"]
         if platform.system() == "Linux":
             data.update(_linux_distro())
         elif platform.system() == "Darwin":
@@ -385,7 +396,6 @@ class NagiosRunnerPlugin(MonitorPlugin):
 
     async def _collect_metrics(self) -> Dict[str, Any]:
         results: Dict[str, Any] = {}
-        worst = 0
         for cmd_cfg in self.commands:
             name = cmd_cfg.get("name")
             command = cmd_cfg.get("command")
@@ -396,10 +406,6 @@ class NagiosRunnerPlugin(MonitorPlugin):
             results[f"{name}_status_code"] = rc
             results[f"{name}_output"] = msg
             results.update({f"{name}_{k}": v for k, v in perf.items()})
-            worst = max(worst, rc)
-        results["overall_status"] = _NAGIOS_STATUS.get(worst, "UNKNOWN")
-        results["overall_status_code"] = worst
-        results["plugin_count"] = len(self.commands)
         return results
 
 
@@ -484,6 +490,12 @@ class CPUMonitorPlugin(MonitorPlugin):
         except Exception:
             pass
 
+        try:
+            with open("/proc/uptime") as fh:
+                data["uptime_seconds"] = int(float(fh.read().split()[0]))
+        except Exception:
+            pass
+
         return data
 
 
@@ -532,6 +544,20 @@ class MemoryMonitorPlugin(MonitorPlugin):
         total = mi.get("MemTotal", 0)
         avail = mi.get("MemAvailable", mi.get("MemFree", 0))
         free = mi.get("MemFree", 0)
+
+        # ZFS ARC is reclaimable but not included in MemAvailable; add it.
+        arc_kb = 0
+        try:
+            with open("/proc/spl/kstat/zfs/arcstats") as _f:
+                for _line in _f:
+                    _p = _line.split()
+                    if len(_p) >= 3 and _p[0] == "size":
+                        arc_kb = int(_p[2]) // 1024
+                        break
+        except (OSError, ValueError):
+            pass
+
+        avail = min(avail + arc_kb, total)
         used = total - avail
         data: Dict[str, Any] = {
             "memory_total": total * 1024,
@@ -698,7 +724,9 @@ async def _load_plugins(cfg: Dict[str, Any]) -> List[Plugin]:
     plugins_cfg: Dict[str, Any] = cfg.get("plugins", {})
     loaded: List[Plugin] = []
     for cls in _ALL_PLUGIN_CLASSES:
-        plugin_cfg = plugins_cfg.get(cls.name) or cfg.get(cls.name, {})
+        plugin_cfg = dict(plugins_cfg.get(cls.name) or cfg.get(cls.name) or {})
+        if "owner" in cfg and "owner" not in plugin_cfg:
+            plugin_cfg["owner"] = cfg["owner"]
         plugin: Plugin = cls(config=plugin_cfg)
         try:
             ok = await plugin.initialize()
@@ -768,7 +796,7 @@ class _HeartbeatProtocol(asyncio.DatagramProtocol):
             msg_id = msg.get("ID")
             now = time.time()
             if msg_id == "ACK":
-                self._conn._handle_ack(now)
+                self._conn._handle_ack(msg, now)
             elif msg_id == "CMD":
                 asyncio.create_task(_handle_command(self._conn, msg))
             elif msg_id == "UPD":
@@ -779,8 +807,7 @@ class _HeartbeatProtocol(asyncio.DatagramProtocol):
             self._log.error("datagram error: %s", e)
 
     def error_received(self, exc):
-        self._log.warning("protocol error on %s: %s — dropping connection", self._conn.addr, exc)
+        self._log.warning("protocol error on %s: %s — will retry", self._conn.addr, exc)
-        self._conn._dead = True
         self._conn.close()
 
 
@@ -796,6 +823,7 @@ class AsyncConnection:
         self.rtts: List[float] = [0.0]
         self._transport: Optional[asyncio.DatagramTransport] = None
         self._dead = False
+        self._request_info: asyncio.Event = asyncio.Event()
         self._log = logging.getLogger(f"hbc.conn.{addr}")
 
     async def open(self) -> bool:
@@ -814,12 +842,14 @@ class AsyncConnection:
             self._transport.close()
             self._transport = None
 
-    def _handle_ack(self, now: float):
+    def _handle_ack(self, msg: Dict[str, Any], now: float):
         rtt = (now - self.lastsend) * 1000.0
         self.rtts.append(rtt)
         if len(self.rtts) > 10:
             self.rtts.pop(0)
         self.ackcount += 1
+        if msg.get("request_update"):
+            self._request_info.set()
 
     async def sendto(self, msg: Dict[str, Any], msg_id: str = "HTB"):
         if self._dead:
@@ -845,6 +875,10 @@ async def _handle_command(conn: AsyncConnection, msg: Dict[str, Any]):
     if not cmd:
         return
     log = logging.getLogger("hbc.cmd")
+    if not _allow_remote_command:
+        log.warning("refused command (allow_remote_command is false): %s", cmd)
+        await conn.sendto({"service": "command", "msg": "Refused: allow_remote_command is false"})
+        return
     log.info("exec: %s", cmd)
     try:
         out = subprocess.check_output(
@@ -875,7 +909,7 @@ async def _handle_update(conn: AsyncConnection):
     log.info("running installer: %s", installer)
     try:
         proc = await asyncio.create_subprocess_exec(
-            installer, "miniclient",
+            installer, "mini",
             stdout=asyncio.subprocess.PIPE,
             stderr=asyncio.subprocess.STDOUT,
         )
@@ -930,7 +964,7 @@ async def _run_info_plugins(conn: AsyncConnection, plugins: List[Plugin]):
         try:
             data = await plugin.collect()
             if data:
-                await conn.sendto({"plugin": plugin.name, **data}, "PLG")
+                await conn.sendto({"plugin": plugin.name, **data, "_interval": plugin.interval}, "PLG")
                 log.info("sent %s", plugin.name)
         except Exception as e:
             log.error("%s collect: %s", plugin.name, e)
@@ -943,7 +977,7 @@ async def _run_monitor_group(conn: AsyncConnection, plugins: List[Plugin], inter
             try:
                 data = await plugin.collect()
                 if data:
-                    await conn.sendto({"plugin": plugin.name, **data}, "PLG")
+                    await conn.sendto({"plugin": plugin.name, **data, "_interval": plugin.interval}, "PLG")
                     log.debug("sent %s", plugin.name)
             except asyncio.CancelledError:
                 raise
@@ -952,6 +986,19 @@ async def _run_monitor_group(conn: AsyncConnection, plugins: List[Plugin], inter
         await _sleep(interval)
 
 
+async def _info_refresh_loop(conn: AsyncConnection, info: List[Plugin]):
+    log = logging.getLogger("hbc.plugins")
+    while _running:
+        await conn._request_info.wait()
+        if not _running:
+            break
+        conn._request_info.clear()
+        log.info("refreshing InfoPlugins on server request")
+        for plugin in info:
+            plugin._cache = None
+        await _run_info_plugins(conn, info)
+
+
 async def _plugin_collector(conn: AsyncConnection, plugins: List[Plugin]):
     info = [p for p in plugins if isinstance(p, InfoPlugin)]
     monitor = [p for p in plugins if isinstance(p, MonitorPlugin)]
@@ -962,12 +1009,10 @@ async def _plugin_collector(conn: AsyncConnection, plugins: List[Plugin]):
     for p in monitor:
         by_interval[p.interval].append(p)
 
-    if by_interval:
+    tasks = [asyncio.create_task(_info_refresh_loop(conn, info))]
-        await asyncio.gather(
+    tasks += [asyncio.create_task(_run_monitor_group(conn, grp, iv))
-            *[asyncio.create_task(_run_monitor_group(conn, grp, iv))
+              for iv, grp in by_interval.items()]
-              for iv, grp in by_interval.items()],
+    await asyncio.gather(*tasks, return_exceptions=True)
-            return_exceptions=True,
-        )
 
 
 # ---------------------------------------------------------------------------
@@ -1011,7 +1056,7 @@ def _reconfigure_syslog(level: int):
 # ---------------------------------------------------------------------------
 
 async def _async_main(args, cfg: Dict[str, Any]) -> int:
-    global _running, _shutdown_event, _active_tasks
+    global _running, _shutdown_event, _active_tasks, send_shutdown
     _running = True
     _shutdown_event = asyncio.Event()
     _active_tasks = []
@@ -1021,36 +1066,47 @@ async def _async_main(args, cfg: Dict[str, Any]) -> int:
     port = cfg.get("hb_port", PORT)
     interval = cfg.get("interval", INTERVAL)
 
-    log.info("starting: %s -> %s port=%d interval=%ds", iam, args.hosts, port, interval)
+    log.info("hbc_mini %s on %s -> %s port=%d interval=%ds",__version__, iam, args.hosts, port, interval)
+
+    af_filter = (socket.AF_INET if getattr(args, "ipv4_only", False)
+                 else socket.AF_INET6 if getattr(args, "ipv6_only", False)
+                 else 0)
 
     connections: List[AsyncConnection] = []
     conn_id = 1
-    for host in args.hosts:
+    _retry_delay = 5
-        try:
+    while _running and not connections:
-            addrs = socket.getaddrinfo(host, port, 0, 0, socket.SOL_UDP)
+        for host in args.hosts:
-        except socket.gaierror as e:
+            try:
-            log.error("cannot resolve %s: %s", host, e)
+                addrs = socket.getaddrinfo(host, port, af_filter, 0, socket.SOL_UDP)
-            continue
+            except socket.gaierror as e:
-        for ai in addrs:
+                log.warning("cannot resolve %s: %s — retrying in %ds", host, e, _retry_delay)
-            conn = AsyncConnection(conn_id, ai[4][0], port, ai[0], iam)
+                continue
-            if await conn.open():
+            for ai in addrs:
-                connections.append(conn)
+                conn = AsyncConnection(conn_id, ai[4][0], port, ai[0], iam)
-                conn_id += 1
+                if await conn.open():
+                    connections.append(conn)
+                    conn_id += 1
+        if not connections:
+            await _sleep(_retry_delay)
+            _retry_delay = min(_retry_delay * 2, 60)
 
     if not connections:
-        log.error("no connections established")
         return 1
 
     # Boot / one-shot message
+    send_shutdown = False
     if args.boot or args.message:
         bmsg: Dict[str, Any] = {"acks": 0}
         if args.boot:
             bmsg["boot"] = 1
+            args.boot = False  # don't repeat on restart
+            send_shutdown = True
         if args.message:
             bmsg["service"] = "service"
             bmsg["msg"] = args.message
-        for c in connections:
+        target = next((c for c in connections if c._transport), connections[0])
-            await c.sendto(bmsg)
+        await target.sendto(bmsg)
         if args.message and not args.daemon:
             await asyncio.sleep(0.3)
             for c in connections:
@@ -1063,6 +1119,13 @@ async def _async_main(args, cfg: Dict[str, Any]) -> int:
     for sig in (signal.SIGTERM, signal.SIGINT):
         loop.add_signal_handler(sig, _stop)
 
+    def _sighup():
+        global _dorestart
+        _dorestart = True
+        _stop()
+
+    loop.add_signal_handler(signal.SIGHUP, _sighup)
+
     for conn in connections:
         _active_tasks.append(asyncio.create_task(_heartbeat_sender(conn, interval)))
 
@@ -1075,11 +1138,13 @@ async def _async_main(args, cfg: Dict[str, Any]) -> int:
         pass
 
     log.info("shutting down")
-    for conn in connections:
+    target = next((c for c in connections if c._transport), connections[0] if connections else None)
+    if target and send_shutdown:
         try:
-            await conn.sendto({"shutdown": 1, "acks": conn.ackcount})
+            await target.sendto({"shutdown": 1, "acks": target.ackcount})
         except Exception:
             pass
+    for conn in connections:
         conn.close()
     await asyncio.sleep(0.3)
     for plugin in plugins:
@@ -1105,6 +1170,9 @@ def main(argv=None):
     parser.add_argument("-d", "--daemon", action="store_true", help="Run as daemon")
     parser.add_argument("-v", "--verbose", action="store_true", help="Verbose output")
     parser.add_argument("-x", "--debug", action="count", default=0, help="Debug level")
+    af_group = parser.add_mutually_exclusive_group()
+    af_group.add_argument("-4", dest="ipv4_only", action="store_true", help="Use IPv4 only")
+    af_group.add_argument("-6", dest="ipv6_only", action="store_true", help="Use IPv6 only")
     parser.add_argument("hosts", nargs="+", help="HBD server(s)")
     args = parser.parse_args(argv)
 
@@ -1121,6 +1189,11 @@ def main(argv=None):
 
     cfg = _load_config(args.configfile)
 
+    global _allow_remote_command
+    _allow_remote_command = bool(cfg.get("allow_remote_command", False))
+    if _allow_remote_command:
+        logging.warning("allow_remote_command is true — CMD packets from the server will be executed")
+
     if args.daemon:
         _daemonize()
         _reconfigure_syslog(level)

scripts/hbc_windows.spec

@@ -0,0 +1,49 @@
+# PyInstaller spec for hbc_windows.exe
+# Build with: pyinstaller hbc_windows.spec
+#
+# Requirements (on Windows):
+#   pip install pyinstaller
+
+block_cipher = None
+
+a = Analysis(
+    ['hbc_windows.py'],
+    pathex=[],
+    binaries=[],
+    datas=[],
+    hiddenimports=[],
+    hookspath=[],
+    hooksconfig={},
+    runtime_hooks=[],
+    excludes=['tkinter', 'unittest', 'email', 'html', 'http', 'urllib', 'xml'],
+    win_no_prefer_redirects=False,
+    win_private_assemblies=False,
+    cipher=block_cipher,
+    noarchive=False,
+)
+
+pyz = PYZ(a.pure, a.zlib_archive, cipher=block_cipher)
+
+exe = EXE(
+    pyz,
+    a.scripts,
+    a.binaries,
+    a.zipfiles,
+    a.datas,
+    [],
+    name='hbc_windows',
+    debug=False,
+    bootloader_ignore_signals=False,
+    strip=False,
+    upx=False,
+    upx_exclude=[],
+    runtime_tmpdir=None,
+    console=True,
+    disable_windowed_traceback=False,
+    argv_emulation=False,
+    target_arch=None,
+    codesign_identity=None,
+    entitlements_file=None,
+    icon=None,
+    version=None,
+)

scripts/install_hbc_windows.ps1

@@ -0,0 +1,126 @@
+#Requires -RunAsAdministrator
+<#
+.SYNOPSIS
+    Install hbc_windows.exe as a Windows Service using NSSM.
+
+.DESCRIPTION
+    Installs the HeartBeat Client as a Windows Service that starts automatically.
+    Requires NSSM (Non-Sucking Service Manager) in PATH or alongside this script.
+    Requires hbc_windows.exe built via: pyinstaller hbc_windows.spec
+
+.PARAMETER Server
+    HBD server hostname or IP address (required).
+
+.PARAMETER ExePath
+    Path to hbc_windows.exe. Defaults to the directory containing this script.
+
+.PARAMETER ServiceName
+    Windows service name. Default: heartbeat-client
+
+.PARAMETER ConfigFile
+    Path to hbc.json config file. Optional.
+
+.PARAMETER LogFile
+    Path to log file. Default: C:\ProgramData\heartbeat\hbc.log
+
+.PARAMETER Interval
+    Heartbeat interval in seconds. Default: 10
+
+.EXAMPLE
+    .\install_hbc_windows.ps1 -Server hbd.example.com
+    .\install_hbc_windows.ps1 -Server hbd.example.com -ConfigFile C:\ProgramData\heartbeat\hbc.json
+#>
+
+param(
+    [Parameter(Mandatory = $true)]
+    [string]$Server,
+
+    [string]$ExePath = "",
+    [string]$ServiceName = "heartbeat-client",
+    [string]$ConfigFile = "",
+    [string]$LogFile = "C:\ProgramData\heartbeat\hbc.log",
+    [int]$Interval = 10
+)
+
+Set-StrictMode -Version Latest
+$ErrorActionPreference = "Stop"
+
+# Locate hbc_windows.exe
+if ($ExePath -eq "") {
+    $ExePath = Join-Path $PSScriptRoot "hbc_windows.exe"
+}
+if (-not (Test-Path $ExePath)) {
+    Write-Error "hbc_windows.exe not found at: $ExePath`nBuild it first with: pyinstaller hbc_windows.spec"
+    exit 1
+}
+
+# Locate NSSM
+$nssm = Get-Command nssm -ErrorAction SilentlyContinue
+if (-not $nssm) {
+    $nssmLocal = Join-Path $PSScriptRoot "nssm.exe"
+    if (Test-Path $nssmLocal) {
+        $nssm = $nssmLocal
+    } else {
+        Write-Error "nssm.exe not found in PATH or alongside this script.`nDownload from https://nssm.cc/download"
+        exit 1
+    }
+} else {
+    $nssm = $nssm.Source
+}
+
+# Build argument list
+$args_list = "--daemon $Server"
+if ($ConfigFile -ne "") {
+    $args_list = "--daemon -c `"$ConfigFile`" $Server"
+}
+if ($LogFile -ne "") {
+    $args_list = "$args_list --log-file `"$LogFile`""
+}
+
+# Create data directory
+$dataDir = "C:\ProgramData\heartbeat"
+if (-not (Test-Path $dataDir)) {
+    New-Item -ItemType Directory -Path $dataDir | Out-Null
+    Write-Host "Created $dataDir"
+}
+
+# Remove existing service if present
+$existing = Get-Service -Name $ServiceName -ErrorAction SilentlyContinue
+if ($existing) {
+    Write-Host "Removing existing service '$ServiceName'..."
+    & $nssm stop $ServiceName 2>$null
+    & $nssm remove $ServiceName confirm
+}
+
+# Install service
+Write-Host "Installing service '$ServiceName'..."
+& $nssm install $ServiceName $ExePath $args_list
+if ($LASTEXITCODE -ne 0) {
+    Write-Error "nssm install failed (exit $LASTEXITCODE)"
+    exit 1
+}
+
+# Configure service
+& $nssm set $ServiceName DisplayName "HeartBeat Client"
+& $nssm set $ServiceName Description "Sends heartbeat and plugin metrics to the HBD monitoring server."
+& $nssm set $ServiceName Start SERVICE_AUTO_START
+& $nssm set $ServiceName AppStdout (Join-Path $dataDir "nssm_stdout.log")
+& $nssm set $ServiceName AppStderr (Join-Path $dataDir "nssm_stderr.log")
+& $nssm set $ServiceName AppRotateFiles 1
+& $nssm set $ServiceName AppRotateBytes 5242880
+
+# Start service
+Write-Host "Starting service '$ServiceName'..."
+& $nssm start $ServiceName
+if ($LASTEXITCODE -ne 0) {
+    Write-Warning "Service installed but failed to start — check logs in $dataDir"
+} else {
+    Write-Host "Service '$ServiceName' started successfully."
+    Write-Host "Log file: $LogFile"
+    Write-Host ""
+    Write-Host "Useful commands:"
+    Write-Host "  nssm status $ServiceName"
+    Write-Host "  nssm stop $ServiceName"
+    Write-Host "  nssm restart $ServiceName"
+    Write-Host "  nssm remove $ServiceName confirm"
+}

scripts/test_nagios.py

@@ -68,8 +68,7 @@ async def test_nagios_runner():
     print(f"   ✓ Collected {len(data)} data points")
     
     print(f"\n4. Results:")
-    print(f"   Overall Status: {data.get('overall_status')} (code: {data.get('overall_status_code')})")
+    print(f"   Data points collected: {len(data)}")
-    print(f"   Plugins Executed: {data.get('plugin_count')}")
     
     # Show individual plugin results
     print(f"\n5. Individual Plugin Results:")

tests/test_client_command.py

@@ -0,0 +1,103 @@
+"""Tests for the allow_remote_command gate on CMD packets in hbc."""
+import asyncio
+import subprocess
+
+import pytest
+
+from hbd.client import main as client_main
+from hbd.client.config import CLIENT_DEFAULTS, load_config
+
+
+class FakeConn:
+    """Stand-in for AsyncConnection that records what the client sends back."""
+
+    def __init__(self):
+        self.sent = []
+
+    async def sendto(self, msg, msg_id="HTB"):
+        self.sent.append(msg)
+
+
+@pytest.fixture
+def conn(monkeypatch):
+    """A fake connection with remote commands disabled and subprocess booby-trapped."""
+    monkeypatch.setattr(client_main, "allow_remote_command", False)
+
+    def explode(*a, **kw):
+        raise AssertionError("subprocess must not run when the command is refused")
+
+    monkeypatch.setattr(subprocess, "check_output", explode)
+    return FakeConn()
+
+
+def test_default_is_false():
+    assert CLIENT_DEFAULTS["allow_remote_command"] is False
+    assert load_config("/nonexistent/hbc.yaml")["allow_remote_command"] is False
+    assert client_main.allow_remote_command is False
+
+
+def test_refused_when_disabled(conn):
+    asyncio.run(client_main.handle_command(conn, {"cmd": "id"}))
+    assert conn.sent == [{
+        "service": "command",
+        "msg": "Refused: allow_remote_command is false",
+    }]
+
+
+def test_refusal_reported_for_every_attempt(conn):
+    for _ in range(3):
+        asyncio.run(client_main.handle_command(conn, {"cmd": "rm -rf /"}))
+    assert len(conn.sent) == 3
+
+
+def test_empty_command_is_ignored_either_way(conn, monkeypatch):
+    asyncio.run(client_main.handle_command(conn, {}))
+    monkeypatch.setattr(client_main, "allow_remote_command", True)
+    asyncio.run(client_main.handle_command(conn, {"cmd": ""}))
+    assert conn.sent == []
+
+
+def test_executed_when_enabled(conn, monkeypatch):
+    monkeypatch.setattr(client_main, "allow_remote_command", True)
+    monkeypatch.setattr(subprocess, "check_output", lambda *a, **kw: b"uid=0(root)")
+    asyncio.run(client_main.handle_command(conn, {"cmd": "id"}))
+    assert conn.sent == [{"service": "command", "msg": "OK uid=0(root)"}]
+
+
+def test_failure_reported_when_enabled(conn, monkeypatch):
+    monkeypatch.setattr(client_main, "allow_remote_command", True)
+
+    def fail(*a, **kw):
+        raise subprocess.CalledProcessError(1, "false")
+
+    monkeypatch.setattr(subprocess, "check_output", fail)
+    asyncio.run(client_main.handle_command(conn, {"cmd": "false"}))
+    assert len(conn.sent) == 1
+    assert conn.sent[0]["msg"].startswith("CalledProcessError ")
+
+
+def test_config_file_can_opt_in(tmp_path):
+    cfg_file = tmp_path / "hbc.yaml"
+    cfg_file.write_text("allow_remote_command: true\n", encoding="utf-8")
+    assert load_config(str(cfg_file))["allow_remote_command"] is True
+
+
+@pytest.mark.parametrize("configured, expected", [(True, True), (False, False), (None, False)])
+def test_async_main_sets_the_flag_from_config(monkeypatch, configured, expected):
+    """async_main must apply the config value before any CMD packet can arrive."""
+    monkeypatch.setattr(client_main, "allow_remote_command", not expected)
+
+    class Bail(Exception):
+        pass
+
+    # gethostname() is the first call after the flag is assigned — bail there so the
+    # test never opens a socket.
+    def bail():
+        raise Bail
+
+    monkeypatch.setattr(client_main.socket, "gethostname", bail)
+
+    cfg = {} if configured is None else {"allow_remote_command": configured}
+    with pytest.raises(Bail):
+        asyncio.run(client_main.async_main(object(), cfg))
+    assert client_main.allow_remote_command is expected

tests/test_config_access.py

@@ -0,0 +1,247 @@
+"""Tests for ownership rules and scoped config merges (hbd.server.config_access)."""
+import pytest
+from hbd.server import config_access as ca
+
+
+# ---------------------------------------------------------------------------
+# Visibility helpers
+# ---------------------------------------------------------------------------
+
+def test_is_global_when_no_owner():
+    assert ca.is_global({"type": "pushover"})
+    assert ca.is_global({"type": "email", "owner": ""})
+    assert ca.is_global(None)          # non-dict is treated as global
+    assert not ca.is_global({"type": "email", "owner": "alice"})
+
+
+def test_user_can_use_global_or_own():
+    assert ca.user_can_use({"type": "pushover"}, "alice")
+    assert ca.user_can_use({"owner": "alice"}, "alice")
+    assert not ca.user_can_use({"owner": "bob"}, "alice")
+
+
+HOSTS = {
+    "web1": {"owner": "alice", "watch": True},
+    "web2": {"owner": "bob", "managers": ["alice"]},
+    "web3": {"owner": "bob", "managers": "carol"},      # string manager form
+    "web4": {"watch": True},                            # unowned
+}
+
+
+def test_user_hosts_owner_and_manager():
+    assert set(ca.user_hosts(HOSTS, "alice")) == {"web1", "web2"}
+    assert set(ca.user_hosts(HOSTS, "carol")) == {"web3"}
+    assert ca.user_hosts(HOSTS, "dave") == {}
+    assert ca.user_hosts(None, "alice") == {}
+
+
+CHANNELS = {
+    "global_ch": {"type": "pushover"},
+    "alice_ch": {"type": "email", "owner": "alice"},
+    "bob_ch": {"type": "signal", "owner": "bob"},
+}
+
+
+def test_user_channels_global_plus_own():
+    assert set(ca.user_channels(CHANNELS, "alice")) == {"global_ch", "alice_ch"}
+    assert set(ca.user_channels(CHANNELS, "dave")) == {"global_ch"}
+
+
+TCS = {
+    "default": {"thresholds": {}},
+    "alice_tc": {"owner": "alice", "thresholds": {}},
+}
+
+
+def test_user_threshold_configs_global_plus_own():
+    assert set(ca.user_threshold_configs(TCS, "alice")) == {"default", "alice_tc"}
+    assert set(ca.user_threshold_configs(TCS, "bob")) == {"default"}
+
+
+# ---------------------------------------------------------------------------
+# merge_hosts_scoped
+# ---------------------------------------------------------------------------
+
+MERGE_HOSTS = {
+    "mine": {"owner": "alice", "watch": True, "notification_channels": ["global_ch"]},
+    "managed": {"owner": "bob", "managers": ["alice"], "watch": True,
+                "notification_channels": ["bob_ch"]},
+    "foreign": {"owner": "bob", "watch": False},
+}
+MERGE_CHANNELS = {
+    "global_ch": {"type": "pushover"},
+    "alice_ch": {"type": "email", "owner": "alice"},
+    "bob_ch": {"type": "signal", "owner": "bob"},
+}
+MERGE_TCS = {"alice_tc": {"owner": "alice"}, "bob_tc": {"owner": "bob"}}
+
+
+def _merge_hosts(payload, existing=None):
+    return ca.merge_hosts_scoped(
+        existing if existing is not None else dict(MERGE_HOSTS),
+        payload, "alice", MERGE_CHANNELS, MERGE_TCS,
+    )
+
+
+def _full_payload(**overrides):
+    """Payload covering alice's full visible subset, with per-host overrides."""
+    p = {
+        "mine": {"owner": "alice", "watch": True, "notification_channels": ["global_ch"]},
+        "managed": {"watch": True, "notification_channels": ["bob_ch"]},
+    }
+    p.update(overrides)
+    return p
+
+
+def test_merge_hosts_preserves_foreign_hosts():
+    result = _merge_hosts(_full_payload())
+    assert result["foreign"] == {"owner": "bob", "watch": False}
+
+
+def test_merge_hosts_manager_edits_settings():
+    result = _merge_hosts(_full_payload(managed={"watch": False, "dyndns": True,
+                                                 "notification_channels": ["bob_ch"]}))
+    assert result["managed"]["watch"] is False
+    assert result["managed"]["dyndns"] is True
+    # access fields carried over untouched
+    assert result["managed"]["owner"] == "bob"
+    assert result["managed"]["managers"] == ["alice"]
+
+
+def test_merge_hosts_manager_cannot_change_owner():
+    with pytest.raises(ca.ScopedMergeError, match="managed"):
+        _merge_hosts(_full_payload(managed={"watch": True, "owner": "alice"}))
+
+
+def test_merge_hosts_manager_cannot_change_managers():
+    with pytest.raises(ca.ScopedMergeError, match="managers"):
+        _merge_hosts(_full_payload(managed={"watch": True, "managers": ["alice", "carol"]}))
+
+
+def test_merge_hosts_manager_same_access_values_ok():
+    # Submitting unchanged access fields is not a violation.
+    result = _merge_hosts(_full_payload(managed={"watch": True, "owner": "bob",
+                                                 "managers": ["alice"],
+                                                 "notification_channels": ["bob_ch"]}))
+    assert result["managed"]["owner"] == "bob"
+
+
+def test_merge_hosts_manager_cannot_delete():
+    payload = {"mine": {"owner": "alice", "watch": True}}  # 'managed' missing
+    with pytest.raises(ca.ScopedMergeError, match="managed"):
+        _merge_hosts(payload)
+
+
+def test_merge_hosts_owner_can_delete():
+    payload = _full_payload()
+    del payload["mine"]
+    result = _merge_hosts(payload)
+    assert "mine" not in result
+    assert "managed" in result and "foreign" in result
+
+
+def test_merge_hosts_owner_can_transfer_ownership():
+    result = _merge_hosts(_full_payload(mine={"owner": "bob", "watch": True}))
+    assert result["mine"]["owner"] == "bob"
+
+
+def test_merge_hosts_new_host_owner_forced():
+    result = _merge_hosts(_full_payload(newhost={"watch": True, "owner": "bob"}))
+    assert result["newhost"]["owner"] == "alice"
+
+
+def test_merge_hosts_cannot_touch_foreign_host():
+    with pytest.raises(ca.ScopedMergeError, match="foreign"):
+        _merge_hosts(_full_payload(foreign={"watch": True}))
+
+
+def test_merge_hosts_cannot_add_foreign_private_channel():
+    with pytest.raises(ca.ScopedMergeError, match="bob_ch"):
+        _merge_hosts(_full_payload(mine={"owner": "alice", "watch": True,
+                                         "notification_channels": ["bob_ch"]}))
+
+
+def test_merge_hosts_keeps_preexisting_foreign_assignment():
+    # 'managed' already has bob_ch; keeping it is fine.
+    result = _merge_hosts(_full_payload())
+    assert result["managed"]["notification_channels"] == ["bob_ch"]
+
+
+def test_merge_hosts_can_add_global_and_own_channel():
+    result = _merge_hosts(_full_payload(mine={"owner": "alice", "watch": True,
+                                              "notification_channels": ["global_ch", "alice_ch"]}))
+    assert result["mine"]["notification_channels"] == ["global_ch", "alice_ch"]
+
+
+def test_merge_hosts_threshold_assignment_validated():
+    with pytest.raises(ca.ScopedMergeError, match="bob_tc"):
+        _merge_hosts(_full_payload(mine={"owner": "alice", "watch": True,
+                                         "threshold_config": ["bob_tc"]}))
+    result = _merge_hosts(_full_payload(mine={"owner": "alice", "watch": True,
+                                              "threshold_config": ["default", "alice_tc"]}))
+    assert result["mine"]["threshold_config"] == ["default", "alice_tc"]
+
+
+def test_merge_hosts_rejects_unknown_fields():
+    with pytest.raises(ca.ScopedMergeError, match="sneaky"):
+        _merge_hosts(_full_payload(mine={"owner": "alice", "watch": True, "sneaky": 1}))
+
+
+def test_merge_hosts_owner_clearing_list_removes_key():
+    result = _merge_hosts(_full_payload(mine={"owner": "alice", "watch": True}))
+    assert "notification_channels" not in result["mine"]
+
+
+# ---------------------------------------------------------------------------
+# merge_threshold_configs_scoped
+# ---------------------------------------------------------------------------
+
+MERGE_EXISTING_TCS = {
+    "default": {"thresholds": {"cpu": {"load": {"warning": 2}}}},
+    "alice_tc": {"owner": "alice", "thresholds": {"cpu": {"load": {"warning": 3}}}},
+    "bob_tc": {"owner": "bob", "thresholds": {}},
+    "global_tc": {"thresholds": {}},
+}
+
+
+def _merge_tcs(payload):
+    return ca.merge_threshold_configs_scoped(dict(MERGE_EXISTING_TCS), payload, "alice")
+
+
+def test_merge_tcs_new_config_owner_forced():
+    result = _merge_tcs({"alice_tc": {"thresholds": {}},
+                         "new_tc": {"thresholds": {}, "owner": "bob"}})
+    assert result["new_tc"]["owner"] == "alice"
+
+
+def test_merge_tcs_edit_own():
+    result = _merge_tcs({"alice_tc": {"thresholds": {"mem": {"used": {"warning": 90}}}}})
+    assert result["alice_tc"]["thresholds"] == {"mem": {"used": {"warning": 90}}}
+    assert result["alice_tc"]["owner"] == "alice"
+
+
+def test_merge_tcs_delete_own_when_missing():
+    result = _merge_tcs({})
+    assert "alice_tc" not in result
+
+
+def test_merge_tcs_preserves_global_and_foreign():
+    result = _merge_tcs({"alice_tc": {"thresholds": {}}})
+    assert result["default"] == MERGE_EXISTING_TCS["default"]
+    assert result["bob_tc"] == MERGE_EXISTING_TCS["bob_tc"]
+    assert result["global_tc"] == MERGE_EXISTING_TCS["global_tc"]
+
+
+def test_merge_tcs_rejects_default():
+    with pytest.raises(ca.ScopedMergeError, match="default"):
+        _merge_tcs({"alice_tc": {"thresholds": {}}, "default": {"thresholds": {}}})
+
+
+def test_merge_tcs_rejects_global_name_collision():
+    with pytest.raises(ca.ScopedMergeError, match="global_tc"):
+        _merge_tcs({"alice_tc": {"thresholds": {}}, "global_tc": {"thresholds": {}}})
+
+
+def test_merge_tcs_rejects_foreign_owned():
+    with pytest.raises(ca.ScopedMergeError, match="bob_tc"):
+        _merge_tcs({"alice_tc": {"thresholds": {}}, "bob_tc": {"thresholds": {}}})

tests/test_configio.py

@@ -0,0 +1,162 @@
+import glob
+import os
+import pytest
+from hbd.server import configio
+
+SAMPLE_YAML = """\
+# Server configuration
+hbd_port: 50004  # HTTP API port
+interval: 20
+users:
+  alice:
+    full_name: Alice Smith
+    admin: true
+notification_channels:
+  pushover_ops:
+    type: pushover
+    token: abc123
+"""
+
+
+def test_read_roundtrip_loads_values(tmp_path):
+    f = tmp_path / ".hb.yaml"
+    f.write_text(SAMPLE_YAML)
+    data = configio.read_roundtrip(str(f))
+    assert data["hbd_port"] == 50004
+    assert data["interval"] == 20
+    assert data["users"]["alice"]["full_name"] == "Alice Smith"
+
+
+def test_write_config_creates_backup(tmp_path):
+    f = tmp_path / ".hb.yaml"
+    f.write_text(SAMPLE_YAML)
+    data = configio.read_roundtrip(str(f))
+    data["interval"] = 30
+    configio.write_config(str(f), data)
+    backups = configio.list_backups(str(f))
+    assert len(backups) == 1
+    assert ".bak." in backups[0]
+
+
+def test_write_config_preserves_comments(tmp_path):
+    f = tmp_path / ".hb.yaml"
+    f.write_text(SAMPLE_YAML)
+    data = configio.read_roundtrip(str(f))
+    data["interval"] = 30
+    configio.write_config(str(f), data)
+    content = f.read_text()
+    assert "# Server configuration" in content
+    assert "# HTTP API port" in content
+
+
+def test_write_config_atomically_replaces_file(tmp_path):
+    f = tmp_path / ".hb.yaml"
+    f.write_text(SAMPLE_YAML)
+    data = configio.read_roundtrip(str(f))
+    data["interval"] = 99
+    configio.write_config(str(f), data)
+    assert not (tmp_path / ".hb.yaml.tmp").exists()
+    data2 = configio.read_roundtrip(str(f))
+    assert data2["interval"] == 99
+
+
+def test_write_config_backup_rotation(tmp_path):
+    cfg = tmp_path / ".hb.yaml"
+    cfg.write_text(SAMPLE_YAML)
+    # Pre-create 10 existing backups with old timestamps
+    for i in range(10):
+        (tmp_path / f".hb.yaml.bak.20260101-{i:06d}").write_text("old")
+    data = configio.read_roundtrip(str(cfg))
+    configio.write_config(str(cfg), data)
+    backups = configio.list_backups(str(cfg))
+    assert len(backups) == 10
+    assert not (tmp_path / ".hb.yaml.bak.20260101-000000").exists()
+
+
+def test_list_backups_newest_first(tmp_path):
+    cfg = tmp_path / ".hb.yaml"
+    cfg.write_text(SAMPLE_YAML)
+    for i in range(3):
+        (tmp_path / f".hb.yaml.bak.20260101-{i:02d}0000").write_text("b")
+    backups = configio.list_backups(str(cfg))
+    assert len(backups) == 3
+    assert backups == sorted(backups, reverse=True)
+
+
+def test_apply_structured_section_server_updates_keys(tmp_path):
+    f = tmp_path / ".hb.yaml"
+    f.write_text(SAMPLE_YAML)
+    data = configio.read_roundtrip(str(f))
+    configio.apply_structured_section(data, "server", {"interval": 60, "hbd_port": 8080})
+    assert data["interval"] == 60
+    assert data["hbd_port"] == 8080
+
+
+def test_apply_structured_section_server_ignores_unknown_keys(tmp_path):
+    f = tmp_path / ".hb.yaml"
+    f.write_text(SAMPLE_YAML)
+    data = configio.read_roundtrip(str(f))
+    configio.apply_structured_section(data, "server", {"interval": 60, "not_a_key": "x"})
+    assert "not_a_key" not in data
+
+
+def test_apply_structured_section_users_replaces_dict(tmp_path):
+    f = tmp_path / ".hb.yaml"
+    f.write_text(SAMPLE_YAML)
+    data = configio.read_roundtrip(str(f))
+    new_users = {"bob": {"full_name": "Bob Jones", "admin": False}}
+    configio.apply_structured_section(data, "users", new_users)
+    assert "alice" not in data["users"]
+    assert data["users"]["bob"]["full_name"] == "Bob Jones"
+
+
+def test_apply_yaml_section_notification_channels(tmp_path):
+    f = tmp_path / ".hb.yaml"
+    f.write_text(SAMPLE_YAML)
+    data = configio.read_roundtrip(str(f))
+    new_yaml = "email_ops:\n  type: email\n  recipients: [ops@example.com]\n"
+    configio.apply_yaml_section(data, "notification_channels", new_yaml)
+    assert "email_ops" in data["notification_channels"]
+    assert "pushover_ops" not in data["notification_channels"]
+
+
+def test_apply_yaml_section_thresholds_maps_to_threshold_configs(tmp_path):
+    f = tmp_path / ".hb.yaml"
+    f.write_text(SAMPLE_YAML)
+    data = configio.read_roundtrip(str(f))
+    configio.apply_yaml_section(data, "thresholds", "default:\n  cpu: 80\n")
+    assert "threshold_configs" in data
+    assert data["threshold_configs"]["default"]["cpu"] == 80
+
+
+def test_apply_yaml_section_dns_replaces_each_key(tmp_path):
+    f = tmp_path / ".hb.yaml"
+    f.write_text(SAMPLE_YAML)
+    data = configio.read_roundtrip(str(f))
+    configio.apply_yaml_section(
+        data, "dns",
+        "nsupdate_bin: /usr/bin/nsupdate\ndyndomains: [dyn.example.com]\n"
+    )
+    assert data["nsupdate_bin"] == "/usr/bin/nsupdate"
+    assert data["dyndomains"] == ["dyn.example.com"]
+
+
+def test_apply_yaml_section_unknown_raises(tmp_path):
+    f = tmp_path / ".hb.yaml"
+    f.write_text(SAMPLE_YAML)
+    data = configio.read_roundtrip(str(f))
+    with pytest.raises(ValueError, match="Unknown YAML section"):
+        configio.apply_yaml_section(data, "nope", "x: 1\n")
+
+
+def test_apply_structured_section_unknown_raises(tmp_path):
+    f = tmp_path / ".hb.yaml"
+    f.write_text(SAMPLE_YAML)
+    data = configio.read_roundtrip(str(f))
+    with pytest.raises(ValueError, match="Unknown structured section"):
+        configio.apply_structured_section(data, "nope", {"x": 1})
+
+
+def test_read_roundtrip_missing_file_raises(tmp_path):
+    with pytest.raises(FileNotFoundError):
+        configio.read_roundtrip(str(tmp_path / "nonexistent.yaml"))

tests/test_events_journal.py

@@ -0,0 +1,265 @@
+"""Tests for the dedicated events journal (write path and read path)."""
+import asyncio
+import json
+
+from hbd.server import journal
+
+
+def _make_journal(tmp_path, **overrides):
+    cfg = {"journal_dir": str(tmp_path), "journal_file": "events.journal"}
+    cfg.update(overrides)
+    j = journal.MessageJournal(cfg)
+    assert asyncio.run(j.initialize())
+    return j
+
+
+def _read_lines(tmp_path, name="events.journal"):
+    return (tmp_path / name).read_text(encoding="utf-8").splitlines()
+
+
+EV1 = {"ts": 1000.0, "host": "h1", "level": "INFO", "service": None, "message": "host up"}
+EV2 = {"ts": 2000.0, "host": "h2", "level": "CRITICAL", "service": "cpu", "message": "cpu high"}
+
+
+def test_log_event_writes_one_json_line(tmp_path):
+    j = _make_journal(tmp_path)
+    asyncio.run(j.log_event(EV1))
+    asyncio.run(j.close())
+    lines = _read_lines(tmp_path)
+    assert len(lines) == 1
+    assert json.loads(lines[0]) == EV1
+
+
+def test_log_event_appends_in_order(tmp_path):
+    j = _make_journal(tmp_path)
+    asyncio.run(j.log_event(EV1))
+    asyncio.run(j.log_event(EV2))
+    asyncio.run(j.close())
+    lines = _read_lines(tmp_path)
+    assert [json.loads(ln)["ts"] for ln in lines] == [1000.0, 2000.0]
+
+
+def test_log_event_rotates_at_max_size(tmp_path):
+    # max_size fits one serialized event line (~77 bytes) but not two, so the
+    # second write triggers exactly one rotation
+    j = _make_journal(tmp_path, journal_max_size=120)
+    asyncio.run(j.log_event(EV1))
+    asyncio.run(j.log_event(EV2))
+    asyncio.run(j.close())
+    backups = list(tmp_path.glob("events.journal.*"))
+    assert len(backups) == 1
+    assert json.loads(backups[0].read_text().splitlines()[0]) == EV1
+    assert json.loads(_read_lines(tmp_path)[0]) == EV2
+
+
+def test_log_event_noop_when_disabled(tmp_path):
+    j = journal.MessageJournal(
+        {"journal_dir": str(tmp_path), "journal_file": "events.journal", "journal_enabled": False}
+    )
+    asyncio.run(j.initialize())
+    asyncio.run(j.log_event(EV1))
+    assert not (tmp_path / "events.journal").exists()
+
+
+def test_backfill_seeds_empty_journal(tmp_path):
+    j = _make_journal(tmp_path)
+    asyncio.run(j.backfill([EV1, EV2]))
+    asyncio.run(j.close())
+    lines = _read_lines(tmp_path)
+    assert len(lines) == 2
+    assert json.loads(lines[0]) == EV1
+
+
+def test_backfill_skipped_when_journal_nonempty(tmp_path):
+    (tmp_path / "events.journal").write_text(json.dumps(EV1) + "\n")
+    j = _make_journal(tmp_path)  # initialize() picks up the existing size
+    asyncio.run(j.backfill([EV2]))
+    asyncio.run(j.close())
+    assert len(_read_lines(tmp_path)) == 1
+
+
+def test_get_events_journal_uses_events_config_keys(tmp_path):
+    journal._events_journal_instance = None
+    try:
+        ej = journal.get_events_journal(
+            {
+                "journal_dir": str(tmp_path),
+                "events_journal_file": "ev.jsonl",
+                "events_journal_max_size": 12345,
+                "events_journal_max_backups": 3,
+            }
+        )
+        assert ej.journal_file == "ev.jsonl"
+        assert ej.max_size == 12345
+        assert ej.max_backups == 3
+        assert ej.journal_dir == tmp_path
+        # singleton: second call returns the same instance
+        assert journal.get_events_journal() is ej
+    finally:
+        journal._events_journal_instance = None
+
+
+def test_get_events_journal_defaults(tmp_path):
+    journal._events_journal_instance = None
+    try:
+        ej = journal.get_events_journal({"journal_dir": str(tmp_path)})
+        assert ej.journal_file == "events.journal"
+        assert ej.max_size == 10 * 1024 * 1024
+        assert ej.max_backups == 10
+    finally:
+        journal._events_journal_instance = None
+
+
+# ---- read path -------------------------------------------------------------
+
+
+def _write_journal(path, events):
+    path.write_text("".join(json.dumps(e) + "\n" for e in events), encoding="utf-8")
+
+
+def _evts(*ts_list):
+    return [
+        {"ts": float(t), "host": f"host{i}", "level": "INFO", "service": None, "message": f"msg {t}"}
+        for i, t in enumerate(ts_list)
+    ]
+
+
+def test_read_events_newest_first(tmp_path):
+    _write_journal(tmp_path / "events.journal", _evts(1, 2, 3))
+    events, more = journal.read_events(tmp_path)
+    assert [e["ts"] for e in events] == [3.0, 2.0, 1.0]
+    assert more is False
+
+
+def test_read_events_limit_and_more(tmp_path):
+    _write_journal(tmp_path / "events.journal", _evts(1, 2, 3))
+    events, more = journal.read_events(tmp_path, limit=2)
+    assert [e["ts"] for e in events] == [3.0, 2.0]
+    assert more is True
+
+
+def test_read_events_before_cursor(tmp_path):
+    _write_journal(tmp_path / "events.journal", _evts(1, 2, 3))
+    events, _ = journal.read_events(tmp_path, before=3.0)
+    assert [e["ts"] for e in events] == [2.0, 1.0]
+
+
+def test_read_events_spans_rotated_files(tmp_path):
+    # rotated backup holds the oldest events; current file the newest
+    _write_journal(tmp_path / "events.journal.20260101-000000", _evts(1, 2))
+    _write_journal(tmp_path / "events.journal.20260201-000000", _evts(3, 4))
+    _write_journal(tmp_path / "events.journal", _evts(5, 6))
+    events, more = journal.read_events(tmp_path, limit=10)
+    assert [e["ts"] for e in events] == [6.0, 5.0, 4.0, 3.0, 2.0, 1.0]
+    assert more is False
+
+
+def test_read_events_pagination_across_files(tmp_path):
+    _write_journal(tmp_path / "events.journal.20260101-000000", _evts(1, 2))
+    _write_journal(tmp_path / "events.journal", _evts(3, 4))
+    page1, more1 = journal.read_events(tmp_path, limit=3)
+    assert [e["ts"] for e in page1] == [4.0, 3.0, 2.0]
+    assert more1 is True
+    page2, more2 = journal.read_events(tmp_path, limit=3, before=page1[-1]["ts"])
+    assert [e["ts"] for e in page2] == [1.0]
+    assert more2 is False
+
+
+def test_read_events_host_filter_substring_case_insensitive(tmp_path):
+    evs = [
+        {"ts": 1.0, "host": "Wentworth", "level": "INFO", "service": None, "message": "a"},
+        {"ts": 2.0, "host": "winter", "level": "INFO", "service": None, "message": "b"},
+    ]
+    _write_journal(tmp_path / "events.journal", evs)
+    events, _ = journal.read_events(tmp_path, host="went")
+    assert [e["host"] for e in events] == ["Wentworth"]
+
+
+def test_read_events_level_filter_exact_case_insensitive(tmp_path):
+    evs = [
+        {"ts": 1.0, "host": "h", "level": "CRITICAL", "service": None, "message": "a"},
+        {"ts": 2.0, "host": "h", "level": "INFO", "service": None, "message": "b"},
+    ]
+    _write_journal(tmp_path / "events.journal", evs)
+    events, _ = journal.read_events(tmp_path, level="critical")
+    assert [e["level"] for e in events] == ["CRITICAL"]
+
+
+def test_read_events_message_filter(tmp_path):
+    evs = [
+        {"ts": 1.0, "host": "h", "level": "INFO", "service": None, "message": "disk almost full"},
+        {"ts": 2.0, "host": "h", "level": "INFO", "service": None, "message": "all quiet"},
+    ]
+    _write_journal(tmp_path / "events.journal", evs)
+    events, _ = journal.read_events(tmp_path, q="Disk")
+    assert [e["ts"] for e in events] == [1.0]
+
+
+def test_read_events_skips_malformed_lines(tmp_path):
+    p = tmp_path / "events.journal"
+    p.write_text('{"ts": 1.0, "host": "h", "level": "INFO", "message": "ok"}\nnot json\n[1,2]\n')
+    events, _ = journal.read_events(tmp_path)
+    assert [e["ts"] for e in events] == [1.0]
+
+
+def test_read_events_predicate(tmp_path):
+    _write_journal(tmp_path / "events.journal", _evts(1, 2))
+    events, _ = journal.read_events(tmp_path, predicate=lambda e: e["host"] == "host0")
+    assert [e["host"] for e in events] == ["host0"]
+
+
+def test_read_events_missing_dir(tmp_path):
+    events, more = journal.read_events(tmp_path / "nope")
+    assert events == [] and more is False
+
+
+def test_filter_events_over_in_memory_ring():
+    ring = _evts(1, 2, 3)  # oldest-first, like data.msgs
+    events, more = journal.filter_events(reversed(ring), limit=2)
+    assert [e["ts"] for e in events] == [3.0, 2.0]
+    assert more is True
+
+
+def test_filter_events_skips_non_dict_entries():
+    ring = [
+        {"ts": 1.0, "host": "h", "level": "INFO", "service": None, "message": "a"},
+        "legacy string entry",
+        None,
+        {"ts": 2.0, "host": "h", "level": "INFO", "service": None, "message": "b"},
+    ]
+    events, more = journal.filter_events(reversed(ring), limit=10)
+    assert [e["ts"] for e in events] == [2.0, 1.0]
+    assert more is False
+
+
+# ---- eventlog wiring --------------------------------------------------------
+
+
+def test_eventlog_writes_to_events_journal(tmp_path):
+    from hbd.server import data, notify
+
+    journal._events_journal_instance = None
+    saved_msgs = data.msgs
+    data.msgs = []
+    try:
+        ej = journal.get_events_journal({"journal_dir": str(tmp_path)})
+
+        async def scenario():
+            await ej.initialize()
+            notify.setup({}, loop=asyncio.get_running_loop())
+            notify.eventlog("h1", "INFO", "hello world")
+            await asyncio.sleep(0.05)  # let the scheduled journal write run
+            await ej.close()
+
+        asyncio.run(scenario())
+        lines = _read_lines(tmp_path)
+        assert len(lines) == 1
+        ev = json.loads(lines[0])
+        assert ev["host"] == "h1"
+        assert ev["level"] == "INFO"
+        assert ev["message"] == "hello world"
+        assert isinstance(ev["ts"], float)
+    finally:
+        journal._events_journal_instance = None
+        data.msgs = saved_msgs
+        notify._loop = None

tests/test_flap.py

@@ -0,0 +1,197 @@
+"""Tests for flap detection (hbd.server.flap) and its hook in notify.send_notification."""
+import asyncio
+import json
+
+import pytest
+
+from hbd.server import flap, hbdclass, notify, users as users_mod
+
+
+@pytest.fixture(autouse=True)
+def clean_flap_state():
+    """Every test starts with an empty state and a 3-in-10-minutes config."""
+    flap._state.clear()
+    flap.setup({"flap_count": 3, "flap_interval": 10})
+    yield
+    flap._state.clear()
+    flap.setup({})
+
+
+def alerts(n, host="h1", service="cpu", level="CRITICAL"):
+    return [flap.observe(host, service, level) for _ in range(n)]
+
+
+def advance(seconds, host="h1", service="cpu"):
+    """Backdate a key's recorded timestamps to simulate the clock moving on."""
+    st = flap._state[(host, service)]
+    st["events"] = [t - seconds for t in st["events"]]
+    if st["ok_since"] is not None:
+        st["ok_since"] -= seconds
+
+
+# --- tripping ---------------------------------------------------------------
+
+def test_alerts_up_to_flap_count_pass():
+    assert alerts(3) == [flap.PASS] * 3
+    assert flap.flapping_services("h1") == []
+
+
+def test_exceeding_flap_count_trips_then_suppresses():
+    assert alerts(5) == [flap.PASS, flap.PASS, flap.PASS, flap.TRIP, flap.SUPPRESS]
+    assert flap.flapping_services("h1") == ["cpu"]
+
+
+def test_warning_and_critical_both_count():
+    flap.observe("h1", "cpu", "WARNING")
+    flap.observe("h1", "cpu", "CRITICAL")
+    flap.observe("h1", "cpu", "WARNING")
+    assert flap.observe("h1", "cpu", "CRITICAL") == flap.TRIP
+
+
+def test_alerts_outside_the_window_do_not_count():
+    alerts(3)
+    advance(11 * 60)  # older than flap_interval
+    assert alerts(3) == [flap.PASS] * 3
+    assert flap.flapping_services("h1") == []
+
+
+def test_recover_alone_never_trips():
+    assert [flap.observe("h1", "cpu", "RECOVER") for _ in range(5)] == [flap.PASS] * 5
+
+
+# --- suppression while flapping --------------------------------------------
+
+def test_recover_and_info_are_suppressed_while_flapping():
+    alerts(4)
+    assert flap.observe("h1", "cpu", "RECOVER") == flap.SUPPRESS
+    assert flap.observe("h1", "cpu", "INFO") == flap.SUPPRESS
+
+
+def test_info_passes_when_not_flapping():
+    assert flap.observe("h1", "cpu", "INFO") == flap.PASS
+
+
+# --- clearing ---------------------------------------------------------------
+
+def test_clears_silently_one_interval_after_recover():
+    alerts(4)
+    flap.observe("h1", "cpu", "RECOVER")
+    advance(10 * 60)
+    assert flap.flapping_services("h1") == []
+    assert flap.observe("h1", "cpu", "CRITICAL") == flap.PASS
+
+
+def test_still_flapping_before_the_interval_elapses():
+    alerts(4)
+    flap.observe("h1", "cpu", "RECOVER")
+    advance(9 * 60)
+    assert flap.flapping_services("h1") == ["cpu"]
+    assert flap.observe("h1", "cpu", "CRITICAL") == flap.SUPPRESS
+
+
+def test_alert_during_the_quiet_window_keeps_it_flapping():
+    alerts(4)
+    flap.observe("h1", "cpu", "RECOVER")
+    advance(9 * 60)
+    flap.observe("h1", "cpu", "CRITICAL")  # resets the quiet window
+    advance(2 * 60)
+    assert flap.flapping_services("h1") == ["cpu"]
+
+
+# --- keying -----------------------------------------------------------------
+
+def test_services_and_hosts_are_tracked_independently():
+    alerts(4, service="cpu")
+    assert flap.observe("h1", "disk", "CRITICAL") == flap.PASS
+    assert flap.observe("h2", "cpu", "CRITICAL") == flap.PASS
+    assert flap.flapping_services("h1") == ["cpu"]
+    assert flap.flapping_services("h2") == []
+
+
+def test_host_level_events_use_the_empty_service_key():
+    alerts(4, service="")
+    assert flap.flapping_services("h1") == [""]
+
+
+# --- disabled ---------------------------------------------------------------
+
+@pytest.mark.parametrize("cfg", [
+    {"flap_count": 0, "flap_interval": 10},
+    {"flap_count": 3, "flap_interval": 0},
+    {},
+])
+def test_disabled_never_suppresses(cfg):
+    flap.setup(cfg)
+    assert alerts(20) == [flap.PASS] * 20
+    assert flap.flapping_services("h1") == []
+
+
+# --- integration through the real dispatch path -----------------------------
+
+NOTIFY_CFG = {
+    "flap_count": 3,
+    "flap_interval": 10,
+    "notification_channels": {"ch1": {"type": "pushover", "token": "t", "user": "u"}},
+    "users": {"alice": {"notification_channels": ["ch1"]}},
+}
+
+
+@pytest.fixture
+def delivered(monkeypatch):
+    """Wire up a host, a user and a stub channel driver; collect what gets delivered."""
+    sent = []
+    monkeypatch.setattr(notify, "_config", dict(notify._config))  # restored on teardown
+    notify.setup(NOTIFY_CFG)
+    users_mod.load_users(NOTIFY_CFG)
+    monkeypatch.setitem(notify._DRIVERS, "pushover", lambda cfg, n: sent.append(n.body) or True)
+    host = hbdclass.Host("flaphost")
+    host.watched = True
+    host.owner = "alice"
+    yield sent, host
+    hbdclass.Host.hosts.pop("flaphost", None)
+    users_mod.load_users({})
+
+
+def _notify(level, service, body):
+    n = notify.Notification(title=f"[{level}] flaphost", body=body, level=level, service=service)
+    asyncio.run(notify.send_notification("flaphost", n))
+
+
+def test_marker_on_the_tripping_notification_then_nothing(delivered):
+    sent, host = delivered
+    for i in range(5):
+        _notify("CRITICAL", "cpu", f"cpu = 9{i}")
+    _notify("RECOVER", "cpu", "cpu = 12")
+
+    assert sent == ["cpu = 90", "cpu = 91", "cpu = 92", f"cpu = 93 {flap.FLAP_MARKER}"]
+    assert host.stateinfo()["flapping"] == ["cpu"]
+
+
+def test_a_flapping_service_does_not_silence_another(delivered):
+    sent, host = delivered
+    for i in range(5):
+        _notify("CRITICAL", "cpu", "cpu = 99")
+    sent.clear()
+    _notify("CRITICAL", "disk", "disk = 91")
+    assert sent == ["disk = 91"]
+
+
+def test_notifications_resume_after_the_state_clears(delivered):
+    sent, host = delivered
+    for i in range(5):
+        _notify("CRITICAL", "cpu", "cpu = 99")
+    _notify("RECOVER", "cpu", "cpu = 12")
+    advance(10 * 60, host="flaphost")
+    assert host.stateinfo()["flapping"] == []
+    sent.clear()
+    _notify("CRITICAL", "cpu", "cpu = 99")
+    assert sent == ["cpu = 99"]
+
+
+def test_stateinfo_flapping_survives_json_encoding(delivered):
+    sent, host = delivered
+    for i in range(5):
+        _notify("CRITICAL", "", "IPv4 overdue")   # host-level, empty service key
+    info = host.stateinfo()
+    assert info["flapping"] == [""]
+    assert json.loads(json.dumps(info))["flapping"] == [""]

tests/test_handle_datagram.py

@@ -20,7 +20,7 @@ def test_handle_cmd_sends_command():
     import hbdclass
 
     ctx = {
-        "config": {"watchhosts": [], "dyndnshosts": []},
+        "config": {"watchhosts": []},
         "hbdclass": hbdclass,
         "log": dummy_noop,
         "email": dummy_noop,

tests/test_http_config_api.py

@@ -0,0 +1,198 @@
+"""Tests for the config read/write API helpers in http.py."""
+import pytest
+from hbd.server import http
+
+
+def test_mask_config_for_api_masks_user_passwords():
+    config = {
+        "hbd_port": 50004,
+        "interval": 20,
+        "users": {
+            "alice": {"full_name": "Alice", "admin": True, "password": "pbkdf2:sha256:abc"},
+        },
+        "oauth": {},
+    }
+    result = http._mask_config_for_api(config)
+    assert result["users"]["alice"]["password"] == "•••"
+    assert result["users"]["alice"]["full_name"] == "Alice"
+
+
+def test_mask_config_for_api_masks_oauth_client_secret():
+    config = {
+        "hbd_port": 50004,
+        "interval": 20,
+        "users": {},
+        "oauth": {
+            "gitea": {"type": "gitea", "url": "https://git.example.com",
+                      "client_id": "cid", "client_secret": "verysecret"},
+        },
+    }
+    result = http._mask_config_for_api(config)
+    assert result["oauth"]["gitea"]["client_secret"] == "•••"
+    assert result["oauth"]["gitea"]["client_id"] == "cid"
+
+
+def test_mask_config_for_api_includes_server_keys():
+    config = {"hbd_port": 50004, "interval": 20, "users": {}, "oauth": {}}
+    result = http._mask_config_for_api(config)
+    assert result["server"]["hbd_port"] == 50004
+    assert result["server"]["interval"] == 20
+
+
+def test_mask_config_for_api_no_password_in_users_leaves_no_key():
+    config = {
+        "hbd_port": 50004,
+        "users": {"bob": {"full_name": "Bob", "admin": False}},
+        "oauth": {},
+    }
+    result = http._mask_config_for_api(config)
+    assert "password" not in result["users"]["bob"]
+
+
+# ---- configio integration for write path ----
+
+def test_write_path_applies_server_section(tmp_path):
+    cfg = tmp_path / ".hb.yaml"
+    cfg.write_text("hbd_port: 50004\ninterval: 20\nusers: {}\n")
+    from hbd.server import configio
+    data = configio.read_roundtrip(str(cfg))
+    configio.apply_structured_section(data, "server", {"interval": 60})
+    configio.write_config(str(cfg), data)
+    data2 = configio.read_roundtrip(str(cfg))
+    assert data2["interval"] == 60
+    assert data2["hbd_port"] == 50004  # unchanged
+
+
+def test_write_path_applies_yaml_section(tmp_path):
+    cfg = tmp_path / ".hb.yaml"
+    cfg.write_text(
+        "hbd_port: 50004\nnotification_channels:\n  old_ch:\n    type: email\n"
+    )
+    from hbd.server import configio
+    data = configio.read_roundtrip(str(cfg))
+    configio.apply_yaml_section(data, "notification_channels", "new_ch:\n  type: pushover\n")
+    configio.write_config(str(cfg), data)
+    data2 = configio.read_roundtrip(str(cfg))
+    assert "new_ch" in data2["notification_channels"]
+    assert "old_ch" not in data2["notification_channels"]
+
+
+def test_write_path_hashes_plaintext_password(tmp_path):
+    cfg = tmp_path / ".hb.yaml"
+    cfg.write_text("hbd_port: 50004\nusers:\n  alice:\n    full_name: Alice\n    admin: true\n    password: pbkdf2:sha256:old\n")
+    from hbd.server import configio
+    from hbd.server import users as users_mod
+    data = configio.read_roundtrip(str(cfg))
+    # Simulate what the POST handler does: hash plaintext password
+    new_users = {"alice": {"full_name": "Alice", "admin": True, "password": "newplaintext"}}
+    for username, attrs in new_users.items():
+        pw = attrs.get("password", "")
+        if pw and not pw.startswith("pbkdf2:"):
+            attrs["password"] = users_mod.hash_password(pw)
+    configio.apply_structured_section(data, "users", new_users)
+    configio.write_config(str(cfg), data)
+    data2 = configio.read_roundtrip(str(cfg))
+    assert data2["users"]["alice"]["password"].startswith("pbkdf2:")
+    assert data2["users"]["alice"]["password"] != "newplaintext"
+
+
+def test_rollback_restores_backup(tmp_path):
+    cfg = tmp_path / ".hb.yaml"
+    cfg.write_text("hbd_port: 50004\ninterval: 20\n")
+    from hbd.server import configio
+    # Make a change to create a backup
+    data = configio.read_roundtrip(str(cfg))
+    data["interval"] = 99
+    configio.write_config(str(cfg), data)
+    backups = configio.list_backups(str(cfg))
+    assert len(backups) == 1
+    # Read the backup and write it back (simulating rollback)
+    backup_data = configio.read_roundtrip(backups[0])
+    configio.write_config(str(cfg), backup_data)
+    restored = configio.read_roundtrip(str(cfg))
+    assert restored["interval"] == 20
+
+
+def test_write_path_preserves_masked_password(tmp_path):
+    """The "•••" sentinel must preserve the existing hash, not write "•••" to disk."""
+    cfg = tmp_path / ".hb.yaml"
+    original_hash = "pbkdf2:sha256:original_hash"
+    cfg.write_text(
+        f"hbd_port: 50004\nusers:\n  alice:\n    full_name: Alice\n    admin: true\n    password: {original_hash}\n"
+    )
+    from hbd.server import configio
+    from hbd.server import users as users_mod
+    data = configio.read_roundtrip(str(cfg))
+    # Simulate what api_config_post does when client sends "•••" back
+    existing_users = data.get("users") or {}
+    users_payload = {"alice": {"full_name": "Alice", "admin": True, "password": "•••"}}
+    for username, attrs in users_payload.items():
+        pw = attrs.get("password", "")
+        if pw and pw != "•••" and not pw.startswith("pbkdf2:"):
+            attrs["password"] = users_mod.hash_password(pw)
+        elif not pw or pw == "•••":
+            existing_hash = (existing_users.get(username) or {}).get("password", "")
+            if existing_hash:
+                attrs["password"] = existing_hash
+            else:
+                attrs.pop("password", None)
+    configio.apply_structured_section(data, "users", users_payload)
+    configio.write_config(str(cfg), data)
+    data2 = configio.read_roundtrip(str(cfg))
+    assert data2["users"]["alice"]["password"] == original_hash, (
+        f"Expected original hash preserved, got: {data2['users']['alice']['password']!r}"
+    )
+
+
+def test_write_path_preserves_oauth_client_secret(tmp_path):
+    """The "•••" sentinel for oauth client_secret must preserve the existing secret."""
+    cfg = tmp_path / ".hb.yaml"
+    original_secret = "real_client_secret_value"
+    cfg.write_text(
+        f"hbd_port: 50004\noauth:\n  gitea:\n    type: gitea\n    url: https://git.example.com\n"
+        f"    client_id: cid123\n    client_secret: {original_secret}\n"
+    )
+    from hbd.server import configio
+    data = configio.read_roundtrip(str(cfg))
+    # Simulate what api_config_post does when client sends "•••" back for client_secret
+    existing_oauth = data.get("oauth") or {}
+    new_oauth = {"gitea": {"type": "gitea", "url": "https://git.example.com", "client_id": "cid123", "client_secret": "•••"}}
+    for name, attrs in new_oauth.items():
+        cs = attrs.get("client_secret", "")
+        if not cs or cs == "•••":
+            existing_cs = (existing_oauth.get(name) or {}).get("client_secret", "")
+            if existing_cs:
+                attrs["client_secret"] = existing_cs
+            else:
+                attrs.pop("client_secret", None)
+    data["oauth"] = new_oauth
+    configio.write_config(str(cfg), data)
+    data2 = configio.read_roundtrip(str(cfg))
+    assert data2["oauth"]["gitea"]["client_secret"] == original_secret, (
+        f"Expected original secret preserved, got: {data2['oauth']['gitea']['client_secret']!r}"
+    )
+
+
+# ---- threshold form payload shape ----
+
+def test_build_threshold_configs_new_shape_with_owner():
+    form = {
+        "servers": {
+            "owner": "alice",
+            "metrics": {"cpu_monitor.load_15min": {"operator": ">", "warning": 4.0,
+                                                   "critical": 8.0, "enabled": True}},
+        },
+    }
+    result = http._build_threshold_configs_from_form(form)
+    assert result["servers"]["owner"] == "alice"
+    assert result["servers"]["thresholds"]["cpu_monitor"]["load_15min"]["warning"] == 4.0
+
+
+def test_build_threshold_configs_empty_owner_means_global():
+    form = {"servers": {"owner": "", "metrics": {"rtt": {"warning": 100.0}}}}
+    result = http._build_threshold_configs_from_form(form)
+    assert "owner" not in result["servers"]
+
+
+def test_build_threshold_configs_ignores_entries_without_metrics():
+    assert http._build_threshold_configs_from_form({"bad": {"owner": "x"}}) == {}

tests/test_http_host_info.py

@@ -0,0 +1,202 @@
+"""Tests for _build_host_info helper in http.py."""
+import pytest
+from unittest.mock import MagicMock
+from hbd.server.http import _build_host_info
+
+
+class _FakeConn:
+    def __init__(self, lastbeat):
+        self.lastbeat = lastbeat
+
+
+class _FakeHost:
+    def __init__(self, name="myhost", owner=None, managers=None,
+                 connections=None, os_data=None, plugin_data=None):
+        self.name = name
+        self.owner = owner
+        self.managers = managers or []
+        self.connections = connections or {}
+        self._os_data = os_data
+        self.plugin_data = plugin_data or {}
+
+    def get_latest_plugin_data(self, plugin_name):
+        if plugin_name == "os_info" and self._os_data is not None:
+            return (1234567890.0, self._os_data)
+        return None
+
+
+def test_build_host_info_basic_fields():
+    host = _FakeHost(owner="alice", managers=["bob", "carol"])
+    result = _build_host_info(host)
+    assert result["owner"] == "alice"
+    assert result["managers"] == ["bob", "carol"]
+    assert result["hbc_version"] is None
+    assert result["hbc_type"] is None
+    assert result["last_packet"] is None
+    assert result["thresholds"] is None
+
+
+def test_build_host_info_no_owner():
+    host = _FakeHost()
+    result = _build_host_info(host)
+    assert result["owner"] is None
+    assert result["managers"] == []
+
+
+def test_build_host_info_reads_hbc_from_os_info():
+    host = _FakeHost(os_data={"hbc_version": "5.3.0", "hbc_type": "full"})
+    result = _build_host_info(host)
+    assert result["hbc_version"] == "5.3.0"
+    assert result["hbc_type"] == "full"
+
+
+def test_build_host_info_hbc_none_when_no_os_info():
+    host = _FakeHost(os_data=None)
+    result = _build_host_info(host)
+    assert result["hbc_version"] is None
+    assert result["hbc_type"] is None
+
+
+def test_build_host_info_last_packet_is_max_lastbeat():
+    host = _FakeHost(connections={
+        "IPv4": _FakeConn(1000.0),
+        "IPv6": _FakeConn(2000.0),
+    })
+    result = _build_host_info(host)
+    assert result["last_packet"] == 2000.0
+
+
+def test_build_host_info_last_packet_none_when_no_connections():
+    host = _FakeHost(connections={})
+    result = _build_host_info(host)
+    assert result["last_packet"] is None
+
+
+def test_build_host_info_thresholds_none_without_checker():
+    host = _FakeHost()
+    result = _build_host_info(host, threshold_checker=None)
+    assert result["thresholds"] is None
+
+
+def test_build_host_info_thresholds_sorted_by_metric():
+    from hbd.server.threshold import ThresholdConfig
+    tc_cpu = ThresholdConfig("cpu_monitor.cpu_percent", warning=80.0, critical=95.0)
+    tc_mem = ThresholdConfig("memory_monitor.memory_percent", warning=85.0, critical=98.0)
+
+    checker = MagicMock()
+    checker.get_thresholds_for_host.return_value = {
+        "memory_monitor.memory_percent": tc_mem,
+        "cpu_monitor.cpu_percent": tc_cpu,
+    }
+
+    host = _FakeHost()
+    result = _build_host_info(host, threshold_checker=checker)
+
+    assert result["thresholds"] is not None
+    assert len(result["thresholds"]) == 2
+    assert result["thresholds"][0]["metric"] == "cpu_monitor.cpu_percent"
+    assert result["thresholds"][0]["warning"] == 80.0
+    assert result["thresholds"][0]["critical"] == 95.0
+    assert result["thresholds"][0]["operator"] == ">"
+    assert result["thresholds"][1]["metric"] == "memory_monitor.memory_percent"
+
+
+def test_build_host_info_thresholds_empty_list_when_no_thresholds():
+    checker = MagicMock()
+    checker.get_thresholds_for_host.return_value = {}
+    host = _FakeHost()
+    result = _build_host_info(host, threshold_checker=checker)
+    assert result["thresholds"] == []
+
+
+def test_build_host_info_threshold_null_warning_critical():
+    from hbd.server.threshold import ThresholdConfig
+    tc = ThresholdConfig("rtt.myhost", warning=None, critical=500.0)
+    checker = MagicMock()
+    checker.get_thresholds_for_host.return_value = {"rtt.myhost": tc}
+    host = _FakeHost()
+    result = _build_host_info(host, threshold_checker=checker)
+    assert result["thresholds"][0]["warning"] is None
+    assert result["thresholds"][0]["critical"] == 500.0
+
+
+def test_build_host_info_nagios_operator_serialized():
+    from hbd.server.threshold import ThresholdConfig
+    tc = ThresholdConfig("nagios_runner.check_http", operator="nagios")
+    checker = MagicMock()
+    checker.get_thresholds_for_host.return_value = {"nagios_runner.check_http": tc}
+    host = _FakeHost()
+    result = _build_host_info(host, threshold_checker=checker)
+    assert result["thresholds"][0]["operator"] == "nagios"
+
+
+def test_build_host_info_covers_suffix_matched_metrics():
+    """memory_monitor.percent threshold covers swap_percent via suffix match."""
+    from hbd.server.threshold import ThresholdConfig
+    tc_pct = ThresholdConfig("memory_monitor.percent", warning=85.0, critical=95.0)
+    checker = MagicMock()
+    checker.get_thresholds_for_host.return_value = {"memory_monitor.percent": tc_pct}
+
+    host = _FakeHost(
+        connections={},
+        os_data=None,
+    )
+    # Simulate plugin_data with both percent and swap_percent fields
+    host.plugin_data = {
+        "memory_monitor": [(1234567890.0, {
+            "percent": 80.0,
+            "swap_percent": 25.0,
+            "available_mb": 2000,
+        })]
+    }
+
+    result = _build_host_info(host, threshold_checker=checker)
+    assert result["thresholds"] is not None
+    t = result["thresholds"][0]
+    assert t["metric"] == "memory_monitor.percent"
+    assert t["covers"] == ["memory_monitor.swap_percent"]
+
+
+def test_build_host_info_covers_empty_when_exact_matches_only():
+    """No covers when all plugin fields match their threshold exactly."""
+    from hbd.server.threshold import ThresholdConfig
+    tc_pct = ThresholdConfig("memory_monitor.percent", warning=85.0, critical=95.0)
+    checker = MagicMock()
+    checker.get_thresholds_for_host.return_value = {"memory_monitor.percent": tc_pct}
+
+    host = _FakeHost()
+    host.plugin_data = {
+        "memory_monitor": [(1234567890.0, {"percent": 80.0})]
+    }
+
+    result = _build_host_info(host, threshold_checker=checker)
+    t = result["thresholds"][0]
+    assert t["covers"] == []
+
+
+class _FakeConnFull:
+    def __init__(self, afam, addr, state, rtt, statetime, lastbeat):
+        self.afam = afam
+        self.addr = addr
+        self.state = state
+        self.rtts = [rtt]
+        self.statetime = statetime
+        self.lastbeat = lastbeat
+
+
+def test_build_host_info_includes_connections():
+    host = _FakeHost(connections={
+        "IPv4": _FakeConnFull("IPv4", "10.0.0.5", "up", 14.2, 1000.0, 2000.0),
+        "IPv6": _FakeConnFull("IPv6", "fd00::5", "overdue", 0, 1500.0, 1900.0),
+    })
+    result = _build_host_info(host)
+    conns = {c["family"]: c for c in result["connections"]}
+    assert conns["IPv4"] == {"family": "IPv4", "addr": "10.0.0.5", "state": "up",
+                             "rtt": 14.2, "statetime": 1000.0, "lastbeat": 2000.0}
+    assert conns["IPv6"]["state"] == "overdue"
+
+
+def test_build_host_info_connections_empty():
+    host = _FakeHost()
+    result = _build_host_info(host)
+    assert result["connections"] == []

tests/test_http_users_me.py

@@ -0,0 +1,123 @@
+"""Tests for PUT /api/0/users/me logic."""
+import pytest
+from hbd.server import users as users_mod
+
+
+def test_hash_password_roundtrip():
+    h = users_mod.hash_password("mysecret")
+    assert h.startswith("pbkdf2:sha256:")
+    assert users_mod.authenticate.__doc__ is not None  # module loaded
+
+
+def test_password_change_requires_correct_current(tmp_path):
+    cfg = tmp_path / ".hb.yaml"
+    initial_hash = users_mod.hash_password("oldpass")
+    cfg.write_text(
+        f"hbd_port: 50004\nusers:\n  alice:\n    full_name: Alice\n    admin: true\n    password: {initial_hash}\n"
+    )
+    users_mod.load_users({"users": {"alice": {"full_name": "Alice", "admin": True, "password": initial_hash}}})
+
+    # Correct current password authenticates
+    assert users_mod.authenticate("alice", "oldpass") is not None
+    # Wrong current password does not authenticate
+    assert users_mod.authenticate("alice", "wrongpass") is None
+
+
+def test_put_users_me_writes_new_fields(tmp_path):
+    """Simulate the write path: read config, update user, write back."""
+    initial_hash = users_mod.hash_password("secret")
+    yaml_content = (
+        "hbd_port: 50004\n"
+        f"users:\n  alice:\n    full_name: Old Name\n    admin: true\n    password: {initial_hash}\n"
+    )
+    cfg = tmp_path / ".hb.yaml"
+    cfg.write_text(yaml_content)
+
+    from hbd.server import configio
+    data = configio.read_roundtrip(str(cfg))
+
+    # Simulate handler updating full_name and avatar
+    user_entry = dict(data["users"]["alice"])
+    user_entry["full_name"] = "New Name"
+    user_entry["avatar"] = "/img/alice.png"
+    data["users"]["alice"] = user_entry
+
+    configio.write_config(str(cfg), data)
+    result = configio.read_roundtrip(str(cfg))
+    assert result["users"]["alice"]["full_name"] == "New Name"
+    assert result["users"]["alice"]["avatar"] == "/img/alice.png"
+    assert result["users"]["alice"]["password"] == initial_hash  # unchanged
+
+
+def test_put_users_me_changes_password(tmp_path):
+    initial_hash = users_mod.hash_password("oldpass")
+    cfg = tmp_path / ".hb.yaml"
+    cfg.write_text(
+        f"hbd_port: 50004\nusers:\n  alice:\n    full_name: Alice\n    password: {initial_hash}\n"
+    )
+    from hbd.server import configio
+    data = configio.read_roundtrip(str(cfg))
+
+    new_hash = users_mod.hash_password("newpass")
+    data["users"]["alice"]["password"] = new_hash
+    configio.write_config(str(cfg), data)
+
+    result = configio.read_roundtrip(str(cfg))
+    # Load users from new config and authenticate with new password
+    new_config = {"users": dict(result["users"])}
+    users_mod.load_users(new_config)
+    assert users_mod.authenticate("alice", "newpass") is not None
+    assert users_mod.authenticate("alice", "oldpass") is None
+
+
+def test_put_users_me_notification_channels(tmp_path):
+    cfg = tmp_path / ".hb.yaml"
+    cfg.write_text(
+        "hbd_port: 50004\n"
+        "notification_channels:\n  pushover_ops:\n    type: pushover\n"
+        "users:\n  alice:\n    full_name: Alice\n    notification_channels: []\n"
+    )
+    from hbd.server import configio
+    data = configio.read_roundtrip(str(cfg))
+    data["users"]["alice"]["notification_channels"] = ["pushover_ops"]
+    configio.write_config(str(cfg), data)
+    result = configio.read_roundtrip(str(cfg))
+    assert result["users"]["alice"]["notification_channels"] == ["pushover_ops"]
+
+
+def test_visible_channels_excludes_private_from_others():
+    """Private channels owned by another user must not appear in the visible set."""
+    from hbd.server import settings as settings_mod
+
+    config = {
+        "notification_channels": {
+            "public_ch":  {"type": "pushover", "token": "t", "user": "u"},
+            "alice_priv": {"type": "email", "owner": "alice", "private": True,
+                           "recipients": ["a@b.com"], "sender": "s@b.com", "smtp_server": "s"},
+            "bob_priv":   {"type": "email", "owner": "bob", "private": True,
+                           "recipients": ["b@b.com"], "sender": "s@b.com", "smtp_server": "s"},
+        }
+    }
+
+    class FakeUser:
+        def __init__(self, username, admin=False):
+            self.username = username
+            self.admin = admin
+
+    alice = FakeUser("alice")
+    bob   = FakeUser("bob")
+    admin = FakeUser("admin", admin=True)
+
+    # Simulate _visible_channels_for_user logic (mirrors http.py implementation)
+    def visible(user):
+        all_channels = config.get("notification_channels") or {}
+        if user.admin:
+            return set(all_channels.keys())
+        return {
+            name for name, cfg in all_channels.items()
+            if not cfg.get("private") or cfg.get("owner") == user.username
+        }
+
+    assert visible(alice) == {"public_ch", "alice_priv"}
+    assert visible(bob)   == {"public_ch", "bob_priv"}
+    assert visible(admin) == {"public_ch", "alice_priv", "bob_priv"}

tests/test_notification_channels_api.py

@@ -0,0 +1,170 @@
+"""Tests for notification channel CRUD via configio helpers and visibility logic."""
+import pytest
+from hbd.server import configio, settings as settings_mod
+
+
+SAMPLE_YAML = """\
+hbd_port: 50004
+notification_channels:
+  pushover_ops:
+    type: pushover
+    token: abc123
+    user: usr456
+"""
+
+
+# ---------------------------------------------------------------------------
+# configio helpers
+# ---------------------------------------------------------------------------
+
+def test_apply_channel_adds_new_entry(tmp_path):
+    f = tmp_path / ".hb.yaml"
+    f.write_text(SAMPLE_YAML)
+    data = configio.read_roundtrip(str(f))
+    configio.apply_channel(data, "email_ops", {"type": "email", "recipients": ["ops@example.com"]})
+    assert "email_ops" in data["notification_channels"]
+    assert data["notification_channels"]["email_ops"]["type"] == "email"
+    # Existing channel preserved
+    assert "pushover_ops" in data["notification_channels"]
+
+
+def test_apply_channel_updates_existing(tmp_path):
+    f = tmp_path / ".hb.yaml"
+    f.write_text(SAMPLE_YAML)
+    data = configio.read_roundtrip(str(f))
+    configio.apply_channel(data, "pushover_ops", {"type": "pushover", "token": "new_tok", "user": "new_usr"})
+    assert data["notification_channels"]["pushover_ops"]["token"] == "new_tok"
+
+
+def test_apply_channel_creates_section_if_absent():
+    data = {"hbd_port": 50004}
+    configio.apply_channel(data, "test_ch", {"type": "pushover", "token": "t", "user": "u"})
+    assert "notification_channels" in data
+    assert "test_ch" in data["notification_channels"]
+
+
+def test_delete_channel_removes_entry(tmp_path):
+    f = tmp_path / ".hb.yaml"
+    f.write_text(SAMPLE_YAML)
+    data = configio.read_roundtrip(str(f))
+    configio.delete_channel(data, "pushover_ops")
+    assert "pushover_ops" not in data["notification_channels"]
+
+
+def test_delete_channel_noop_for_missing():
+    data = {"notification_channels": {"ch1": {"type": "pushover"}}}
+    configio.delete_channel(data, "nonexistent")  # must not raise
+    assert "ch1" in data["notification_channels"]
+
+
+def test_delete_channel_noop_when_no_section():
+    data = {}
+    configio.delete_channel(data, "anything")  # must not raise
+
+
+def test_apply_channel_persisted_after_write(tmp_path):
+    f = tmp_path / ".hb.yaml"
+    f.write_text(SAMPLE_YAML)
+    data = configio.read_roundtrip(str(f))
+    configio.apply_channel(data, "signal_ops", {"type": "signal", "user": "+1", "recipient": "+2"})
+    configio.write_config(str(f), data)
+    result = configio.read_roundtrip(str(f))
+    assert "signal_ops" in result["notification_channels"]
+    assert result["notification_channels"]["signal_ops"]["user"] == "+1"
+    # Original channel preserved
+    assert "pushover_ops" in result["notification_channels"]
+
+
+def test_delete_channel_persisted_after_write(tmp_path):
+    f = tmp_path / ".hb.yaml"
+    f.write_text(SAMPLE_YAML)
+    data = configio.read_roundtrip(str(f))
+    configio.delete_channel(data, "pushover_ops")
+    configio.write_config(str(f), data)
+    result = configio.read_roundtrip(str(f))
+    assert "pushover_ops" not in (result.get("notification_channels") or {})
+
+
+# ---------------------------------------------------------------------------
+# Visibility logic (owner-presence rule, hbd.server.config_access)
+# ---------------------------------------------------------------------------
+
+from hbd.server import config_access as ca  # noqa: E402
+
+
+def _visible(config, user):
+    all_channels = config.get("notification_channels") or {}
+    if user.get("admin"):
+        return set(all_channels.keys())
+    return set(ca.user_channels(all_channels, user["username"]))
+
+
+CONFIG_VISIBILITY = {
+    "notification_channels": {
+        "pub_ch":      {"type": "pushover", "token": "t", "user": "u"},
+        "alice_priv":  {"type": "email", "owner": "alice",
+                        "recipients": ["a@a.com"], "sender": "s@a.com", "smtp_server": "s"},
+        "bob_priv":    {"type": "signal", "owner": "bob", "user": "+1", "recipient": "+2"},
+        "stale_flag":  {"type": "pushover", "token": "t2", "user": "u2", "private": True},
+    }
+}
+
+
+def test_global_channel_visible_to_all():
+    for uname in ("alice", "bob", "carol"):
+        assert "pub_ch" in _visible(CONFIG_VISIBILITY, {"username": uname, "admin": False})
+
+
+def test_owned_channel_visible_only_to_owner():
+    alice = {"username": "alice", "admin": False}
+    bob = {"username": "bob", "admin": False}
+    assert "alice_priv" in _visible(CONFIG_VISIBILITY, alice)
+    assert "alice_priv" not in _visible(CONFIG_VISIBILITY, bob)
+    assert "bob_priv" in _visible(CONFIG_VISIBILITY, bob)
+    assert "bob_priv" not in _visible(CONFIG_VISIBILITY, alice)
+
+
+def test_admin_sees_all_channels():
+    admin = {"username": "adminuser", "admin": True}
+    assert _visible(CONFIG_VISIBILITY, admin) == {"pub_ch", "alice_priv", "bob_priv", "stale_flag"}
+
+
+def test_stale_private_flag_without_owner_is_global():
+    """Owner-presence is the single signal; a leftover private flag is ignored."""
+    alice = {"username": "alice", "admin": False}
+    assert "stale_flag" in _visible(CONFIG_VISIBILITY, alice)
+
+
+# ---------------------------------------------------------------------------
+# Channel type schemas
+# ---------------------------------------------------------------------------
+
+def test_all_required_types_in_schema():
+    for t in ("pushover", "email", "signal", "matrix", "sms_voipms"):
+        assert t in settings_mod.CHANNEL_TYPE_SCHEMAS
+
+
+def test_schema_fields_have_required_keys():
+    for type_id, schema in settings_mod.CHANNEL_TYPE_SCHEMAS.items():
+        assert "label" in schema, f"{type_id} missing label"
+        assert "fields" in schema, f"{type_id} missing fields"
+        for f in schema["fields"]:
+            for k in ("key", "label", "type", "required"):
+                assert k in f, f"{type_id} field missing {k!r}"
+
+
+def test_secret_fields_use_secret_type():
+    """Known secret fields must be typed 'secret' so the UI masks them."""
+    secret_keys = {"token", "user_key", "api_key", "api_password",
+                   "smtp_password", "access_token"}
+    for type_id, schema in settings_mod.CHANNEL_TYPE_SCHEMAS.items():
+        for f in schema["fields"]:
+            if f["key"] in secret_keys:
+                assert f["type"] == "secret", (
+                    f"{type_id}.{f['key']} should be type 'secret'"
+                )
+
+
+def test_channel_labels_not_empty():
+    for type_id, schema in settings_mod.CHANNEL_TYPE_SCHEMAS.items():
+        assert schema["label"].strip(), f"{type_id} has empty label"

tests/test_oauth.py

@@ -0,0 +1,602 @@
+import logging
+import time as time_mod
+from unittest.mock import AsyncMock, MagicMock, patch
+from urllib.parse import urlparse, parse_qs
+
+import pytest
+
+from hbd.server import oauth
+from hbd.server import users as users_mod
+from hbd.server.users import User
+
+
+CFG_OFF = {}
+CFG_ON = {
+    "oauth": {
+        "gitea": {
+            "url": "https://git.example.com",
+            "client_id": "cid",
+            "client_secret": "csec",
+        }
+    }
+}
+CFG_PARTIAL = {"oauth": {"gitea": {"url": "https://git.example.com"}}}
+
+
+@pytest.fixture(autouse=True)
+def clear_oauth_states():
+    oauth._states.clear()
+    yield
+    oauth._states.clear()
+
+
+@pytest.fixture(autouse=True)
+def reset_users_dict():
+    original = dict(users_mod.users)
+    yield
+    users_mod.users = original
+
+
+
+def test_make_state_returns_unique_tokens():
+    s1 = oauth.make_state()
+    s2 = oauth.make_state()
+    assert s1 != s2
+    assert len(s1) == 64  # 32 bytes hex
+
+
+def test_validate_state_valid():
+    state = oauth.make_state()
+    assert oauth.validate_state(state) is True
+
+
+def test_validate_state_consumed_on_use():
+    state = oauth.make_state()
+    oauth.validate_state(state)
+    assert oauth.validate_state(state) is False  # replay rejected
+
+
+def test_validate_state_unknown():
+    assert oauth.validate_state("notastate") is False
+
+
+def test_validate_state_expired(monkeypatch):
+    state = oauth.make_state()
+    # Wind expiry into the past
+    monkeypatch.setitem(oauth._states, state, time_mod.time() - 1000)
+    assert oauth.validate_state(state) is False
+
+
+def _reset_users(entries=None):
+    users_mod.users = entries or {}
+
+
+def test_provision_oauth_user_new():
+    _reset_users()
+    user = users_mod.provision_oauth_user("gituser", "Git User", "https://example.com/avatar.png")
+    assert user.username == "gituser"
+    assert user.full_name == "Git User"
+    assert user.avatar == "https://example.com/avatar.png"
+    assert user.admin is False
+    assert user.password_hash == ""
+    assert "gituser" in users_mod.users
+
+
+def test_provision_oauth_user_no_password_login():
+    _reset_users()
+    user = users_mod.provision_oauth_user("gituser", "Git User", "")
+    assert user.check_password("anything") is False
+
+
+def test_provision_oauth_user_existing_updates_profile():
+    existing = User(
+        username="alice",
+        full_name="Old Name",
+        avatar="old.png",
+        password_hash="pbkdf2:sha256:1:salt:abc",
+        admin=True,
+        notification_channels=["chan1"],
+    )
+    _reset_users({"alice": existing})
+    user = users_mod.provision_oauth_user("alice", "New Name", "new.png")
+    assert user.full_name == "New Name"
+    assert user.avatar == "new.png"
+    # Preserved
+    assert user.admin is True
+    assert user.password_hash == "pbkdf2:sha256:1:salt:abc"
+    assert user.notification_channels == ["chan1"]
+
+
+def test_provision_oauth_user_does_not_overwrite_with_empty():
+    existing = User(username="bob", full_name="Bob", avatar="bob.png")
+    _reset_users({"bob": existing})
+    user = users_mod.provision_oauth_user("bob", "", "")
+    assert user.full_name == "Bob"
+    assert user.avatar == "bob.png"
+
+
+def test_provision_oauth_user_survives_config_reload():
+    _reset_users()
+    users_mod.provision_oauth_user("oauthonly", "OAuth Only", "https://example.com/a.png")
+    assert "oauthonly" in users_mod.users
+    # Reload with empty config — OAuth user should survive
+    users_mod.load_users({})
+    assert "oauthonly" in users_mod.users
+
+
+
+# ---------------------------------------------------------------------------
+# Integration-style tests: callback logic chain
+# ---------------------------------------------------------------------------
+
+
+@pytest.mark.asyncio
+async def test_callback_invalid_state_rejects():
+    """Verify validate_state returns False for unknown state tokens."""
+    fake_state = "this-is-not-a-real-state"
+    assert oauth.validate_state(fake_state) is False
+
+
+@pytest.mark.asyncio
+async def test_full_oauth_flow_chain():
+    """Integration-style test: state → exchange → fetch → provision chain."""
+    p = _gitea_provider()
+    redirect_uri = "https://hbd.example.com/login/oauth/gitea/callback"
+
+    state = oauth.make_state()
+    assert oauth.validate_state(state) is True
+
+    mock_token_response = AsyncMock()
+    mock_token_response.status = 200
+    mock_token_response.json = AsyncMock(return_value={"access_token": "flow_token"})
+
+    mock_user_response = AsyncMock()
+    mock_user_response.status = 200
+    mock_user_response.json = AsyncMock(return_value={
+        "login": "flowuser",
+        "full_name": "Flow User",
+        "avatar_url": "https://git.example.com/avatars/flow.png",
+    })
+
+    mock_session = MagicMock()
+    mock_session.post = MagicMock(return_value=AsyncMock(
+        __aenter__=AsyncMock(return_value=mock_token_response),
+        __aexit__=AsyncMock(return_value=False),
+    ))
+    mock_session.get = MagicMock(return_value=AsyncMock(
+        __aenter__=AsyncMock(return_value=mock_user_response),
+        __aexit__=AsyncMock(return_value=False),
+    ))
+
+    with patch("hbd.server.oauth.aiohttp.ClientSession", return_value=AsyncMock(
+        __aenter__=AsyncMock(return_value=mock_session),
+        __aexit__=AsyncMock(return_value=False),
+    )):
+        token = await oauth.exchange_code(p, "authcode", redirect_uri)
+        profile = await oauth.fetch_user(p, token)
+
+    assert token == "flow_token"
+    assert profile["login"] == "flowuser"
+
+    _reset_users()
+    user = users_mod.provision_oauth_user(
+        profile["login"], profile["full_name"], profile["avatar_url"]
+    )
+    assert user.username == "flowuser"
+    assert user.check_password("anything") is False
+
+
+# ---------------------------------------------------------------------------
+# get_providers()
+# ---------------------------------------------------------------------------
+
+CFG_GITHUB = {
+    "oauth": {
+        "github": {"type": "github", "client_id": "ghid", "client_secret": "ghs"},
+    }
+}
+
+CFG_NEXTCLOUD = {
+    "oauth": {
+        "nc": {
+            "type": "nextcloud",
+            "url": "https://nc.example.com",
+            "client_id": "ncid",
+            "client_secret": "ncs",
+        }
+    }
+}
+
+CFG_MULTI = {
+    "oauth": {
+        "mygitea": {
+            "type": "gitea",
+            "url": "https://git.example.com",
+            "client_id": "cid",
+            "client_secret": "cs",
+            "label": "Work Gitea",
+            "logo": "https://example.com/logo.png",
+        },
+        "github": {"type": "github", "client_id": "ghid", "client_secret": "ghs"},
+        "nc": {
+            "type": "nextcloud",
+            "url": "https://nc.example.com",
+            "client_id": "ncid",
+            "client_secret": "ncs",
+        },
+    }
+}
+
+
+def test_get_providers_backward_compat_no_type_field():
+    """Old config without 'type' defaults to gitea."""
+    providers = oauth.get_providers(CFG_ON)
+    assert len(providers) == 1
+    p = providers[0]
+    assert p.name == "gitea"
+    assert p.type == "gitea"
+    assert p.label == "Gitea"
+    assert p.client_id == "cid"
+    assert p.authorize_url == "https://git.example.com/login/oauth/authorize"
+    assert p.token_url == "https://git.example.com/login/oauth/access_token"
+    assert p.profile_url == "https://git.example.com/api/v1/user"
+    assert p.scope == "user:email"
+    assert p.profile_data_path == []
+
+
+def test_get_providers_multiple():
+    providers = oauth.get_providers(CFG_MULTI)
+    assert len(providers) == 3
+    names = [p.name for p in providers]
+    assert "mygitea" in names
+    assert "github" in names
+    assert "nc" in names
+
+
+def test_get_providers_custom_label_and_logo():
+    providers = oauth.get_providers(CFG_MULTI)
+    gitea = next(p for p in providers if p.name == "mygitea")
+    assert gitea.label == "Work Gitea"
+    assert gitea.logo == "https://example.com/logo.png"
+
+
+def test_get_providers_github_default_label():
+    providers = oauth.get_providers(CFG_GITHUB)
+    assert providers[0].label == "GitHub"
+    assert providers[0].logo == ""
+
+
+def test_get_providers_github_fixed_urls():
+    providers = oauth.get_providers(CFG_GITHUB)
+    p = providers[0]
+    assert p.authorize_url == "https://github.com/login/oauth/authorize"
+    assert p.token_url == "https://github.com/login/oauth/access_token"
+    assert p.profile_url == "https://api.github.com/user"
+    assert p.scope == "read:user"
+
+
+def test_get_providers_nextcloud_urls_and_path():
+    providers = oauth.get_providers(CFG_NEXTCLOUD)
+    p = providers[0]
+    assert p.authorize_url == "https://nc.example.com/apps/oauth2/authorize"
+    assert p.token_url == "https://nc.example.com/apps/oauth2/api/v1/token"
+    assert p.profile_url == "https://nc.example.com/ocs/v2.php/cloud/user?format=json"
+    assert p.profile_data_path == ["ocs", "data"]
+    assert p.scope == ""
+
+
+def test_get_providers_skips_missing_client_id(caplog):
+    cfg = {"oauth": {"gitea": {"url": "https://git.example.com", "client_secret": "cs"}}}
+    with caplog.at_level(logging.WARNING, logger="hbd.server.oauth"):
+        result = oauth.get_providers(cfg)
+    assert result == []
+    assert "missing" in caplog.text.lower()
+
+
+def test_get_providers_skips_missing_client_secret(caplog):
+    cfg = {"oauth": {"gitea": {"url": "https://git.example.com", "client_id": "cid"}}}
+    with caplog.at_level(logging.WARNING, logger="hbd.server.oauth"):
+        result = oauth.get_providers(cfg)
+    assert result == []
+    assert "missing" in caplog.text.lower()
+
+
+def test_get_providers_skips_missing_url_for_gitea(caplog):
+    cfg = {"oauth": {"gitea": {"type": "gitea", "client_id": "cid", "client_secret": "cs"}}}
+    with caplog.at_level(logging.WARNING, logger="hbd.server.oauth"):
+        result = oauth.get_providers(cfg)
+    assert result == []
+    assert "url" in caplog.text.lower()
+
+
+def test_get_providers_skips_missing_url_for_nextcloud(caplog):
+    cfg = {"oauth": {"nc": {"type": "nextcloud", "client_id": "cid", "client_secret": "cs"}}}
+    with caplog.at_level(logging.WARNING, logger="hbd.server.oauth"):
+        result = oauth.get_providers(cfg)
+    assert result == []
+    assert "url" in caplog.text.lower()
+
+
+def test_get_providers_github_no_url_required():
+    providers = oauth.get_providers(CFG_GITHUB)
+    assert len(providers) == 1
+
+
+def test_get_providers_skips_unknown_type(caplog):
+    cfg = {"oauth": {"mystery": {"type": "saml", "client_id": "cid", "client_secret": "cs"}}}
+    import logging
+    with caplog.at_level(logging.WARNING, logger="hbd.server.oauth"):
+        result = oauth.get_providers(cfg)
+    assert result == []
+    assert "saml" in caplog.text
+
+
+def test_get_providers_empty_config():
+    assert oauth.get_providers({}) == []
+    assert oauth.get_providers(CFG_OFF) == []
+
+
+# ---------------------------------------------------------------------------
+# build_auth_url / exchange_code / fetch_user (generic, ResolvedProvider-based)
+# ---------------------------------------------------------------------------
+
+def _gitea_provider() -> oauth.ResolvedProvider:
+    return oauth.get_providers(CFG_ON)[0]
+
+
+def _github_provider() -> oauth.ResolvedProvider:
+    return oauth.get_providers(CFG_GITHUB)[0]
+
+
+def _nextcloud_provider() -> oauth.ResolvedProvider:
+    return oauth.get_providers(CFG_NEXTCLOUD)[0]
+
+
+def test_build_auth_url_gitea():
+    p = _gitea_provider()
+    url = oauth.build_auth_url(p, "teststate", "https://hbd.example.com/login/oauth/gitea/callback")
+    parsed = urlparse(url)
+    qs = parse_qs(parsed.query)
+    assert parsed.netloc == "git.example.com"
+    assert parsed.path == "/login/oauth/authorize"
+    assert qs["client_id"] == ["cid"]
+    assert qs["state"] == ["teststate"]
+    assert qs["scope"] == ["user:email"]
+    assert qs["response_type"] == ["code"]
+    assert qs["redirect_uri"] == ["https://hbd.example.com/login/oauth/gitea/callback"]
+
+
+def test_build_auth_url_github():
+    p = _github_provider()
+    url = oauth.build_auth_url(p, "st", "https://hbd.example.com/login/oauth/github/callback")
+    parsed = urlparse(url)
+    qs = parse_qs(parsed.query)
+    assert parsed.netloc == "github.com"
+    assert qs["scope"] == ["read:user"]
+
+
+def test_build_auth_url_nextcloud_no_scope_param():
+    """Nextcloud scope is empty — the 'scope' key must be absent from the URL."""
+    p = _nextcloud_provider()
+    url = oauth.build_auth_url(p, "st", "https://hbd.example.com/login/oauth/nc/callback")
+    qs = parse_qs(urlparse(url).query)
+    assert "scope" not in qs
+
+
+@pytest.mark.asyncio
+async def test_exchange_code_generic_returns_token():
+    p = _gitea_provider()
+    redirect_uri = "https://hbd.example.com/login/oauth/gitea/callback"
+    mock_response = AsyncMock()
+    mock_response.status = 200
+    mock_response.json = AsyncMock(return_value={"access_token": "tok123"})
+
+    mock_session = MagicMock()
+    mock_session.post = MagicMock(return_value=AsyncMock(
+        __aenter__=AsyncMock(return_value=mock_response),
+        __aexit__=AsyncMock(return_value=False),
+    ))
+
+    with patch("hbd.server.oauth.aiohttp.ClientSession", return_value=AsyncMock(
+        __aenter__=AsyncMock(return_value=mock_session),
+        __aexit__=AsyncMock(return_value=False),
+    )):
+        token = await oauth.exchange_code(p, "mycode", redirect_uri)
+    assert token == "tok123"
+
+
+@pytest.mark.asyncio
+async def test_exchange_code_sends_accept_json():
+    """Accept: application/json must be present for all providers (required by GitHub)."""
+    p = _github_provider()
+    captured_headers = {}
+
+    mock_response = AsyncMock()
+    mock_response.status = 200
+    mock_response.json = AsyncMock(return_value={"access_token": "ghtoken"})
+
+    mock_session = MagicMock()
+
+    def capture_post(url, **kwargs):
+        captured_headers.update(kwargs.get("headers", {}))
+        return AsyncMock(
+            __aenter__=AsyncMock(return_value=mock_response),
+            __aexit__=AsyncMock(return_value=False),
+        )
+
+    mock_session.post = capture_post
+
+    with patch("hbd.server.oauth.aiohttp.ClientSession", return_value=AsyncMock(
+        __aenter__=AsyncMock(return_value=mock_session),
+        __aexit__=AsyncMock(return_value=False),
+    )):
+        await oauth.exchange_code(p, "code", "https://hbd.example.com/login/oauth/github/callback")
+
+    assert captured_headers.get("Accept") == "application/json"
+
+
+@pytest.mark.asyncio
+async def test_exchange_code_raises_on_error_status():
+    p = _gitea_provider()
+    mock_response = AsyncMock()
+    mock_response.status = 401
+    mock_response.text = AsyncMock(return_value="unauthorized")
+
+    mock_session = MagicMock()
+    mock_session.post = MagicMock(return_value=AsyncMock(
+        __aenter__=AsyncMock(return_value=mock_response),
+        __aexit__=AsyncMock(return_value=False),
+    ))
+
+    with patch("hbd.server.oauth.aiohttp.ClientSession", return_value=AsyncMock(
+        __aenter__=AsyncMock(return_value=mock_session),
+        __aexit__=AsyncMock(return_value=False),
+    )):
+        with pytest.raises(oauth.OAuthError):
+            await oauth.exchange_code(p, "badcode", "https://hbd.example.com/login/oauth/gitea/callback")
+
+
+@pytest.mark.asyncio
+async def test_exchange_code_raises_when_no_access_token():
+    p = _gitea_provider()
+    mock_response = AsyncMock()
+    mock_response.status = 200
+    mock_response.json = AsyncMock(return_value={"error": "bad_request"})
+
+    mock_session = MagicMock()
+    mock_session.post = MagicMock(return_value=AsyncMock(
+        __aenter__=AsyncMock(return_value=mock_response),
+        __aexit__=AsyncMock(return_value=False),
+    ))
+
+    with patch("hbd.server.oauth.aiohttp.ClientSession", return_value=AsyncMock(
+        __aenter__=AsyncMock(return_value=mock_session),
+        __aexit__=AsyncMock(return_value=False),
+    )):
+        with pytest.raises(oauth.OAuthError):
+            await oauth.exchange_code(p, "mycode", "https://hbd.example.com/login/oauth/gitea/callback")
+
+
+@pytest.mark.asyncio
+async def test_fetch_user_gitea_returns_profile():
+    p = _gitea_provider()
+    mock_response = AsyncMock()
+    mock_response.status = 200
+    mock_response.json = AsyncMock(return_value={
+        "login": "alice",
+        "full_name": "Alice Smith",
+        "avatar_url": "https://git.example.com/avatars/alice.png",
+    })
+
+    mock_session = MagicMock()
+    mock_session.get = MagicMock(return_value=AsyncMock(
+        __aenter__=AsyncMock(return_value=mock_response),
+        __aexit__=AsyncMock(return_value=False),
+    ))
+
+    with patch("hbd.server.oauth.aiohttp.ClientSession", return_value=AsyncMock(
+        __aenter__=AsyncMock(return_value=mock_session),
+        __aexit__=AsyncMock(return_value=False),
+    )):
+        profile = await oauth.fetch_user(p, "tok123")
+
+    assert profile == {
+        "login": "alice",
+        "full_name": "Alice Smith",
+        "avatar_url": "https://git.example.com/avatars/alice.png",
+    }
+
+
+@pytest.mark.asyncio
+async def test_fetch_user_github_maps_name_field():
+    p = _github_provider()
+    mock_response = AsyncMock()
+    mock_response.status = 200
+    mock_response.json = AsyncMock(return_value={
+        "login": "bobgh",
+        "name": "Bob GitHub",
+        "avatar_url": "https://avatars.githubusercontent.com/u/1",
+    })
+
+    mock_session = MagicMock()
+    mock_session.get = MagicMock(return_value=AsyncMock(
+        __aenter__=AsyncMock(return_value=mock_response),
+        __aexit__=AsyncMock(return_value=False),
+    ))
+
+    with patch("hbd.server.oauth.aiohttp.ClientSession", return_value=AsyncMock(
+        __aenter__=AsyncMock(return_value=mock_session),
+        __aexit__=AsyncMock(return_value=False),
+    )):
+        profile = await oauth.fetch_user(p, "ghtoken")
+
+    assert profile["login"] == "bobgh"
+    assert profile["full_name"] == "Bob GitHub"
+    assert profile["avatar_url"] == "https://avatars.githubusercontent.com/u/1"
+
+
+@pytest.mark.asyncio
+async def test_fetch_user_nextcloud_nested_extraction():
+    """Nextcloud profile is nested under ocs.data; avatar is absent."""
+    p = _nextcloud_provider()
+    mock_response = AsyncMock()
+    mock_response.status = 200
+    mock_response.json = AsyncMock(return_value={
+        "ocs": {
+            "meta": {"status": "ok", "statuscode": 200},
+            "data": {
+                "id": "ncuser",
+                "display-name": "NC User",
+                "email": "nc@example.com",
+            },
+        }
+    })
+
+    mock_session = MagicMock()
+    mock_session.get = MagicMock(return_value=AsyncMock(
+        __aenter__=AsyncMock(return_value=mock_response),
+        __aexit__=AsyncMock(return_value=False),
+    ))
+
+    with patch("hbd.server.oauth.aiohttp.ClientSession", return_value=AsyncMock(
+        __aenter__=AsyncMock(return_value=mock_session),
+        __aexit__=AsyncMock(return_value=False),
+    )):
+        profile = await oauth.fetch_user(p, "nctoken")
+
+    assert profile["login"] == "ncuser"
+    assert profile["full_name"] == "NC User"
+    assert profile["avatar_url"] == ""  # Nextcloud has no avatar field
+
+
+@pytest.mark.asyncio
+async def test_fetch_user_raises_on_error_status():
+    p = _gitea_provider()
+    mock_response = AsyncMock()
+    mock_response.status = 401
+    mock_response.text = AsyncMock(return_value="unauthorized")
+
+    mock_session = MagicMock()
+    mock_session.get = MagicMock(return_value=AsyncMock(
+        __aenter__=AsyncMock(return_value=mock_response),
+        __aexit__=AsyncMock(return_value=False),
+    ))
+
+    with patch("hbd.server.oauth.aiohttp.ClientSession", return_value=AsyncMock(
+        __aenter__=AsyncMock(return_value=mock_session),
+        __aexit__=AsyncMock(return_value=False),
+    )):
+        with pytest.raises(oauth.OAuthError):
+            await oauth.fetch_user(p, "badtoken")
+
+
+def test_is_enabled_with_valid_provider():
+    assert oauth.is_enabled(CFG_ON) is True
+
+
+def test_is_enabled_false_when_no_providers():
+    assert oauth.is_enabled(CFG_OFF) is False
+
+
+def test_is_enabled_false_partial_config():
+    assert oauth.is_enabled(CFG_PARTIAL) is False

tests/test_settings_sections.py

@@ -0,0 +1,206 @@
+import pytest
+from hbd.server import settings as settings_mod
+
+CFG = {
+    "hbd_port": 50004,
+    "interval": 20,
+    "grace": 2,
+    "users": {
+        "alice": {"full_name": "Alice Smith", "admin": True, "password": "pbkdf2:sha256:abc",
+                  "notification_channels": ["pushover_ops"]},
+    },
+    "oauth": {
+        "gitea": {"type": "gitea", "url": "https://git.example.com",
+                  "client_id": "cid", "client_secret": "csec", "label": "Sign in with Gitea"},
+    },
+    "notification_channels": {
+        "pushover_ops": {"type": "pushover", "token": "tok", "user": "usr"},
+    },
+    "hosts": {},
+}
+
+
+def test_sections_have_section_mode():
+    sections = settings_mod.get_settings_sections(CFG)
+    for s in sections:
+        assert "section_mode" in s, f"Section {s['id']} missing section_mode"
+        assert s["section_mode"] in ("form", "yaml", "channels", "hosts", "thresholds")
+
+
+def test_sections_have_api_section():
+    sections = settings_mod.get_settings_sections(CFG)
+    for s in sections:
+        assert "api_section" in s, f"Section {s['id']} missing api_section"
+
+
+def test_network_section_has_editable_fields():
+    sections = settings_mod.get_settings_sections(CFG)
+    network = next(s for s in sections if s["id"] == "network")
+    assert network["section_mode"] == "form"
+    assert network["api_section"] == "server"
+    editable = [f for f in network["fields"] if f["editable"]]
+    assert len(editable) >= 2  # hbd_port, ws_port at minimum
+
+
+def test_thresholds_and_dns_section_modes():
+    sections = settings_mod.get_settings_sections(CFG)
+    by_id = {s["id"]: s for s in sections}
+    assert by_id["thresholds"]["section_mode"] == "thresholds"
+    assert by_id["thresholds"]["api_section"] == "thresholds"
+    assert by_id["dns"]["section_mode"] == "form"
+    assert by_id["dns"]["api_section"] == "dns"
+
+
+def test_hosts_section_uses_hosts_mode():
+    sections = settings_mod.get_settings_sections(CFG)
+    hosts_sec = next(s for s in sections if s["id"] == "hosts")
+    assert hosts_sec["section_mode"] == "hosts"
+    assert hosts_sec["api_section"] == "hosts"
+
+
+def test_channels_section_uses_channels_mode():
+    sections = settings_mod.get_settings_sections(CFG)
+    ch_sec = next(s for s in sections if s["id"] == "channels")
+    assert ch_sec["section_mode"] == "channels"
+    assert ch_sec["api_section"] == "notification_channels"
+    assert len(ch_sec["channels"]) == 1
+    ch = ch_sec["channels"][0]
+    assert ch["name"] == "pushover_ops"
+    assert ch["type"] == "pushover"
+    assert "owner" in ch
+    assert "editable" in ch
+
+
+def test_channel_type_schemas_exported():
+    assert hasattr(settings_mod, "CHANNEL_TYPE_SCHEMAS")
+    for required_type in ("pushover", "email", "signal", "matrix", "sms_voipms"):
+        assert required_type in settings_mod.CHANNEL_TYPE_SCHEMAS
+        schema = settings_mod.CHANNEL_TYPE_SCHEMAS[required_type]
+        assert "label" in schema
+        assert "fields" in schema
+        for f in schema["fields"]:
+            assert "key" in f
+            assert "type" in f
+            assert "required" in f
+
+
+def test_oauth_section_exists():
+    sections = settings_mod.get_settings_sections(CFG)
+    oauth = next((s for s in sections if s["id"] == "oauth"), None)
+    assert oauth is not None
+    assert oauth["section_mode"] == "form"
+    assert oauth["api_section"] == "oauth"
+    assert len(oauth["providers"]) == 1
+    assert oauth["providers"][0]["name"] == "gitea"
+    assert oauth["providers"][0]["client_secret"] == "•••"
+
+
+def test_all_channel_names_returned():
+    result = settings_mod.get_settings_data(CFG)
+    assert "all_channel_names" in result
+    assert "pushover_ops" in result["all_channel_names"]
+
+
+def test_users_section_has_user_list():
+    sections = settings_mod.get_settings_sections(CFG)
+    users_sec = next(s for s in sections if s["id"] == "users")
+    assert users_sec["section_mode"] == "form"
+    assert users_sec["api_section"] == "users"
+    assert len(users_sec["users"]) == 1
+    assert users_sec["users"][0]["username"] == "alice"
+    # Password hash never exposed
+    assert "password" not in users_sec["users"][0]
+
+
+# ---------------------------------------------------------------------------
+# Per-user filtering (owners/managers on the settings page)
+# ---------------------------------------------------------------------------
+
+from types import SimpleNamespace  # noqa: E402
+
+
+MULTI_CFG = {
+    **CFG,
+    "users": {
+        "alice": {"full_name": "Alice", "admin": True, "password": "x"},
+        "bob": {"full_name": "Bob", "admin": False, "password": "x"},
+    },
+    "notification_channels": {
+        "global_ch": {"type": "pushover", "token": "t", "user": "u"},
+        "bob_ch": {"type": "pushover", "token": "t", "user": "u", "owner": "bob"},
+        "carol_ch": {"type": "pushover", "token": "t", "user": "u", "owner": "carol"},
+    },
+    "threshold_configs": {
+        "bob_tc": {"owner": "bob", "thresholds": {}},
+        "carol_tc": {"owner": "carol", "thresholds": {}},
+    },
+    "hosts": {
+        "bobhost": {"owner": "bob"},
+        "managedhost": {"owner": "carol", "managers": ["bob"]},
+        "carolhost": {"owner": "carol"},
+    },
+}
+
+BOB = SimpleNamespace(username="bob", admin=False)
+ADMIN = SimpleNamespace(username="alice", admin=True)
+
+
+def test_nonadmin_sees_only_three_sections():
+    sections = settings_mod.get_settings_sections(MULTI_CFG, user=BOB)
+    assert [s["id"] for s in sections] == ["channels", "hosts", "thresholds"]
+
+
+def test_nonadmin_sections_hide_admin_fields():
+    sections = settings_mod.get_settings_sections(MULTI_CFG, user=BOB)
+    for s in sections:
+        assert s["fields"] == []
+
+
+def test_nonadmin_hosts_filtered_with_is_owner():
+    sections = settings_mod.get_settings_sections(MULTI_CFG, user=BOB)
+    hosts = next(s for s in sections if s["id"] == "hosts")["hosts"]
+    by_name = {h["name"]: h for h in hosts}
+    assert set(by_name) == {"bobhost", "managedhost"}
+    assert by_name["bobhost"]["is_owner"] is True
+    assert by_name["managedhost"]["is_owner"] is False
+
+
+def test_nonadmin_channels_filtered_with_editable():
+    sections = settings_mod.get_settings_sections(MULTI_CFG, user=BOB)
+    chans = {c["name"]: c for c in next(s for s in sections if s["id"] == "channels")["channels"]}
+    assert set(chans) == {"global_ch", "bob_ch"}
+    assert chans["bob_ch"]["editable"] is True
+    assert chans["global_ch"]["editable"] is False
+
+
+def test_admin_sees_everything_with_editable():
+    sections = settings_mod.get_settings_sections(MULTI_CFG, user=ADMIN)
+    ids = [s["id"] for s in sections]
+    assert "network" in ids and "users" in ids
+    chans = {c["name"]: c for c in next(s for s in sections if s["id"] == "channels")["channels"]}
+    assert set(chans) == {"global_ch", "bob_ch", "carol_ch"}
+    assert all(c["editable"] for c in chans.values())
+
+
+def test_settings_data_pickers_filtered_for_nonadmin():
+    data = settings_mod.get_settings_data(MULTI_CFG, user=BOB)
+    assert data["all_channel_names"] == ["bob_ch", "global_ch"]
+    assert data["all_threshold_configs"] == ["bob_tc"]
+
+
+def test_no_user_means_admin_view():
+    sections = settings_mod.get_settings_sections(MULTI_CFG)  # auth disabled
+    assert len(sections) > 3
+
+
+def test_filtering_unaffected_by_other_users_in_config():
+    """The users-section loop must not clobber the requesting username (regression)."""
+    cfg = dict(MULTI_CFG)
+    cfg["users"] = {
+        "alice": {"full_name": "Alice", "admin": True, "password": "x"},
+        "bob": {"full_name": "Bob", "admin": False, "password": "x"},
+        "zed": {"full_name": "Zed", "admin": False, "password": "x"},  # bob is not last
+    }
+    sections = settings_mod.get_settings_sections(cfg, user=BOB)
+    hosts = {h["name"] for h in next(s for s in sections if s["id"] == "hosts")["hosts"]}
+    assert hosts == {"bobhost", "managedhost"}

tox.ini

@@ -1,9 +1,8 @@
 [tox]
 envlist = py, lint, mypy
-skipsdist = True
 
 [testenv]
-deps = -rrequirements-dev.txt
+extras = dev
 commands =
     pytest -q