phase: 06_system_page_gpu_verdicts

All criteria verified and the project is fully green.

**Verified (final pass — all 3 phase-06 tasks were already in `complete/`)**
- `classify("gpu.busy_pct", 99)` / `gpu.mem_used_pct` → `None`; `THRESHOLDS` has no `gpu.*` row; `alerts.py`/`api.py` carry no GPU verdict; live browser: deskwork GPU at **91 %** and **98 %** renders `class="hero"`, `color=rgb(22,25,31)` (plain text), no bar, no `.sev-*`.
- GPU temp on the hero `.sub` from live agents: `Radeon 8060S Graphics · VRAM 165.1 MiB / 512.0 MiB · pool 78% · 48 °C`, R9700 `· 40 °C`, gateway prints none (no `n/a`, no `0 °C`).
- Charts collapsed by default: fresh load = 3 panels / 3 toggles / **0 open / 0 `<svg>`**; expand → 3 SVGs (510 px wide), repaints across polls (legend 3.3 %→5.6 %), collapsed siblings keep their own height (803 vs 358/339).
- `gpu.busy_pct@0` still answers from `/api/system/history` (345 points) — verdict gone, data intact.

**Tests / lint / coverage**
- `bash .agents/validate.sh` → **EXIT=0**; `uv run pytest --cov --cov-fail-under=94` → **550 passed, 8 skipped, 97.91 %**; `uv run ruff check src tests agent` → clean; `uv run pyright` → 0 errors; `node --check` all scripts clean; `uv pip list --outdated` empty.
- `uv run pytest -m live --live` → **8 passed** (real agents/engines/upsd); dev server restarted, `/api/health` all sources `ok`.
- New screenshots: `data/logs/phase06-final-collapsed-default.png`, `data/logs/phase06-final-expanded-gpu-busy-uncoloured.png`.

**Completion criteria:** classify→None ✅ · pytest+94 % gate ✅ · ruff+pyright ✅ · validate.sh ✅ · screenshots of the three behaviours on live data ✅ · AGENTS.md severity section claims no GPU rows ✅

**Docs:** fixed 3 items — (1) `AGENTS.md` pytest count stale (546 → 550); (2) `app.js` `severityFor` docstring still claimed "`page-ups.js` still calls it (phase 05 converts it)", contradicting the code/tests/AGENTS.md (comment-only); rest of README/AGENTS reviewed in full (env table vs `config.py`+`.env.example`, agent routes incl. `/schema`, alerts rules, page descriptions, contracts) — consistent.

**Notable:** fixed a real pre-existing defect in `scripts/dev-server.sh` — `stop` slept 1 s and deleted the pidfile without waiting for exit, so `restart` with a browser open overlapped two servers on one SQLite file, producing a `database is locked` store refusal that `/system` rendered as a red event. Now waits (bounded 20 s) then SIGKILLs; verified a restart with live polling logs no such error/event. No live infra config was changed.

**Next pending phase:** `.agents/phases/todo/07_llm_monitor_style_pages`

scripts/dev-server.sh

@@ -39,8 +39,26 @@ start() {
 
 stop() {
   if running; then
-    kill "$(cat "$PIDFILE")" 2>/dev/null
-    sleep 1
+    local pid
+    pid="$(cat "$PIDFILE")"
+    kill "$pid" 2>/dev/null
+    # Wait for the process to actually go away before handing the pidfile back. A bare
+    # `sleep 1` was not enough: uvicorn's graceful shutdown waits for in-flight requests,
+    # so `restart` with a browser open started the new server while the old one still
+    # held the SQLite file — the dying process then logged `database is locked` for a poll
+    # it had already fetched, which the app correctly turned into a red store-error event.
+    # Two servers on one DB is the one way this script can fabricate an incident, so the
+    # wait is bounded rather than optimistic, and SIGKILL is the last resort.
+    local i=0
+    while kill -0 "$pid" 2>/dev/null; do
+      i=$((i + 1))
+      if [ "$i" -gt 60 ]; then
+        echo "pid $pid ignored SIGTERM after 20s; sending SIGKILL"
+        kill -9 "$pid" 2>/dev/null
+        break
+      fi
+      sleep 0.25
+    done
     echo "stopped"
   else
     echo "not running"