[MyNameIsEdi]

Summary

• SQLite run history — replaces per-run JSON files in test-results/runs/ with a single better-sqlite3 database (test-results/runs.db). archiveRun() INSERTs into the DB; /api/playwright/history, /api/playwright/results/:runId, and DELETE /api/playwright/history all query SQLite directly.
• Remove Quick Presets — deleted PRESETS constant, applyPreset(), noPresets prop, and the entire preset card from the Settings panel.
• Run Output above Run History — terminal log panel now appears immediately after the Insights row, above the RunsPanel history table, so live output is the first thing you see after a run.
• Expanded DocsView — grew from 5 to 8 sections; added Authentication, Visual & A11y (axe-core), and CI/CD (GitHub Actions) chapters with code examples.
• Stop button fix — runAbortRef useRef pattern ensures the Stop button aborts the active SSE fetch without racing against subsequent runs.
• History from SQLite only — removed all localStorage caching of run history to prevent stale pre-migration data surfacing on page load.
• streamSummary crash fix — removed stale RUNS_DIR reference left over from JSON-file era; reads raw_json from SQLite when runId is present.

Test plan

• npm install — better-sqlite3 installs cleanly on Node 20
• npm run dev — server starts; test-results/runs.db is created automatically
• Click Run Tests → run appears in /api/playwright/history after completion
• Click a past run in the history table → suite results and log replay correctly
• Click Clear History → history table empties
• Settings panel has no Quick Presets section
• Run Output terminal appears above the Run History table
• Stop button aborts cleanly; clicking Run Tests again works immediately
• npx tsc --noEmit and cd ui && npx tsc --noEmit — zero type errors
• npm run format:check — passes

Summary by CodeRabbit

• New Features

  • Persistent run history and full run logs now stored and replayable from a central archive.
  • Docs view expanded with Authentication, Visual/A11y, and CI/CD sections (runnable examples).

• Changed

  • Run output panel relocated in the dashboard layout; history now queried from the central store (newest-first).
  • Removed Quick Presets and localStorage run-history caching.

• Bug Fixes

  • Stop button reliably aborts live runs; resolved a crash when loading archived run summaries.

• Chores

  • Added SQLite runtime dependency for persisted history.

[coderabbitai]

Caution

Review failed

Pull request was closed or merged during review

📝 Walkthrough

Walkthrough

This PR migrates Playwright run history persistence from individual JSON files under test-results/runs/ to a SQLite database. The server initializes a runs table at startup, the archive function stores normalized test results and capped logs in the database, API endpoints query SQLite instead of the filesystem, and the AI summary helper reads archived snapshots from the database when a runId is provided.

Changes

Run History SQLite Persistence

Layer / File(s)	Summary
Documentation of changes CHANGELOG.md, README.md	CHANGELOG records new Dashboard UI features, immediate history refresh on [DONE], DocsView expansion, and SSE/streamSummary fixes. README updates the Persistent Run History + Logs row to reflect test-results/runs.db.
Dependencies and imports package.json, server/index.ts	Add better-sqlite3 and @types/better-sqlite3; server/index.ts imports Database and Node fs sync helpers.
Database initialization server/index.ts	Create test-results/ if needed, open test-results/runs.db, and create runs table for metadata, normalized counts, capped run_log, and raw_json.
Run archiving via SQLite server/index.ts	archiveRun reads pw-results.json, normalizes results, computes totals, caps run_log, and upserts a runs row containing the full raw_json instead of writing per-run JSON files.
API endpoints for archived runs server/index.ts	GET /api/playwright/history queries runs ordered newest-first (optional limit); GET /api/playwright/results/:runId returns normalized results and run_log from SQLite; DELETE /api/playwright/history clears the table.
Post-run summary integration server/index.ts	streamSummary now prefers raw_json from the runs table when creds.runId is supplied; otherwise it falls back to pw-results.json on disk.
Dashboard UI updates ui/src/pages/PlaywrightDashboard.tsx	Remove localStorage run-history hydrate/persist, clear stale pre-migration key, add runAbortRef and integrate SSE abort handling, change Stop button to abort active SSE, remove Quick Presets and Presets bar, update SettingsView invocation, relocate Run Output panel, and refresh Docs view content.

Sequence Diagram

sequenceDiagram
  participant Dashboard
  participant Server
  participant SQLite
  Dashboard->>Server: GET /api/playwright/history or /results/:runId
  Server->>SQLite: SELECT raw_json, run_log FROM runs WHERE id=?
  SQLite-->>Server: row(s)
  Server-->>Dashboard: normalized results + run_log
  Dashboard->>Server: POST to start run (SSE)
  Server-->>Dashboard: SSE events (stdout, summary_*)
  Dashboard->>Dashboard: runAbortRef aborts SSE controller on Stop

Loading

Estimated code review effort

🎯 4 (Complex) | ⏱️ ~45 minutes

Possibly related PRs

• MyNameIsEdi/open-qa#117: Overlaps in dashboard SSE/live run handling and immediate history refresh after [DONE].
• MyNameIsEdi/open-qa#114: Overlaps in SSE/run-handling and hydration of archived run logs in the dashboard client.
• MyNameIsEdi/open-qa#116: Related client changes for fetching archived runLog and adapting to server-provided immutable run snapshots.

Poem

🐰 I found the logs beneath the bog,

A SQLite burrow for each cog,
No scattered JSON to misplace,
One comfy DB holds every trace,
Hooray — the dashboard hops with grace!

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 62.50% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The PR title clearly and specifically summarizes the main changes: SQLite run history persistence, removal of Quick Presets feature, reordering of run output UI, and expanded documentation—all of which are confirmed in the changeset.
Description check	✅ Passed	The PR description provides a comprehensive summary of all major changes and a detailed test plan, but does not follow the provided template structure with standard sections like Type of change or Checklist.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch feat/playwright-dashboard-upgrade

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}