Restore archived tasks, fix dashboard scroll-reset, bind ornith, add Playwright smoke test

- **Restore 82 completed tasks** from tasks/complete/ back to tasks/ top
  level (all <7 days old per the cleanup policy; premature bulk archive
  was fixed).
- **Dashboard: fix scroll-reset on auto-refresh** — renderBoard rebuilds
  the board via innerHTML every 2s, destroying each column-body's
  scrollTop. Now snapshots column-body scrollTop + board.scrollLeft +
  view.scrollTop before rebuild and restores after (matched by
  PHASE_GROUPS index).
- **Dashboard UI additions** (pre-existing unstaged work): approval
  section cards, transition buttons, inline artifact editor (textarea for
  writing missing SPEC/VERDICT/etc from the detail modal).
- **Bind ornith as Implement model** — config.md: Model explicit to
  omlx/Ornith-1.0-35B-4bit-mlx, context window 32768. Interactive
  autopilot already used ornith via opencode default; now explicit.
- **Fix cleanup stub** — automaton-cleanup.sh had a stale --project arg
  pointing at a pytest temp dir (test isolation leak). Rewired to point
  at ~/.automaton.
- **Fix plist-isolation test** — test asserted host plist doesn't exist,
  but a real install creates it. Now snapshots mtime before run, asserts
  unchanged after (only a write during the test counts as bleed).
- **New Playwright smoke test** (tests/test_dashboard_ui.py) — 2 tests:
  board renders tasks, column scroll survives auto-refresh tick.
  Verified the test fails without the scroll fix (scrollTop resets to 0).
  Skipped via importorskip when playwright is absent (main CI stays
  green).
- **Clarify SI loop scope in README** — new-project onboarding section
  documents the framework-scoped self-improvement loop and options
  (leave/pause/create project loop).
- **CHANGELOG** documents all changes including the known model-divergence
  gap (mde tasks marked complete but per-role model binding was never
  implemented).

tasks/rewrite-vram-detection-python/SPEC.md

@@ -0,0 +1,32 @@
+# SPEC: Rewrite VRAM Detection as Python Script
+
+## Goal
+Replace the fragile `scripts/vram_detect.sh` with a testable Python script that produces the same JSON output.
+
+## Requirements
+1. Create `scripts/vram_detect.py` with the same CLI interface:
+   - `--model` / `-m`
+   - `--project` / `-p`
+   - Optional positional model name
+2. Preserve the human-readable output sections and the `=== JSON Output ===` block.
+3. Fix known bugs:
+   - Undefined `$headroom_pct` in manual mode.
+   - Hardcoded `"headroom": 0.25` in JSON output.
+   - GPU-counting/output ordering bug.
+   - Enforce 10KB file-read limit for API config files.
+4. Delete `scripts/vram_detect.sh`.
+5. Update references in `prompts/orchestrate.md`, `prompts/decompose.md`, `prompts/onboarding.md`, `README.md`, and `config.md` to point to the Python script.
+
+## Acceptance Criteria
+- [ ] `python scripts/vram_detect.py` runs and emits valid JSON.
+- [ ] Manual-mode headroom is read from `config.md`, not hardcoded.
+- [ ] Large API config files are read with a 10KB limit.
+- [ ] `scripts/vram_detect.sh` no longer exists.
+- [ ] All prompts/docs reference the Python script.
+
+## Non-Goals
+- Changing the recommendation algorithm.
+- Adding GPU vendor detection beyond Linux `nvidia-smi`, `lspci`, and `/proc/meminfo`.
+
+## Stop Condition
+When all acceptance criteria are met, output "CONTRACT_MET".