@sun-asterisk/sungen 2.6.1 → 2.6.3

package/src/orchestrator/templates/ai-instructions/github-skill-sungen-dashboard.md

@@ -0,0 +1,121 @@
+---
+name: sungen-dashboard
+description: 'Build a single-file HTML dashboard summarising test cases & Playwright results. Auto-loaded by /sungen:dashboard.'
+user-invocable: false
+---
+
+## Purpose
+
+Generate `qa/dashboard/index.html` — a single-file, share-ready test report covering all (or selected) screens and flows. The HTML embeds the snapshot data inline, so it can be emailed, Slacked, or committed and opened by anyone via `file://` without any server.
+
+**This skill delegates all heavy work to the `sungen dashboard` CLI.** The CLI is the source of truth for parsing logic — do NOT re-parse files in AI. Your role is only to:
+
+1. Invoke the CLI.
+2. Show its output verbatim.
+3. Help the user decide next actions.
+
+---
+
+## Architecture
+
+```
+User → /sungen:dashboard [screen...]
+       │
+       ▼
+  sungen dashboard CLI  (deterministic — no AI tokens)
+  ├─ Discovery: qa/screens/* + qa/flows/*
+  ├─ Reuse delivery parsers (.feature, .spec.ts, test-data.yaml, results.json)
+  ├─ Build DashboardSnapshot JSON
+  ├─ Write qa/dashboard/history/<runId>.json (max 20 retained, oldest pruned)
+  ├─ Inject payload into pre-built HTML template
+  └─ Write qa/dashboard/index.html (~1MB, fully self-contained)
+```
+
+Source modules: `src/dashboard/*.ts` + `src/exporters/json-exporter.ts`.
+UI source: `dashboard/` (built once, ships in npm package as `dist/dashboard/templates/index.html`).
+
+---
+
+## Required sources (CLI tolerates missing files)
+
+| # | Source | Path | Required? |
+|---|--------|------|-----------|
+| 1 | Feature file | `qa/screens/<name>/features/<name>.feature` (or `qa/flows/...`) | Yes — screens without a feature are skipped |
+| 2 | Test data | `qa/screens/<name>/test-data/<name>.yaml` | Optional — `{{vars}}` fall back to literal |
+| 3 | Compiled spec | `specs/generated/<name>/<name>.spec.ts` | Optional — flagged as "Not compiled" if missing |
+| 4 | Test results | `specs/generated/<name>/<name>-test-result.json` (or `test-results/results.json`) | Optional — TCs show as "Pending" if missing |
+
+Unlike `sungen delivery`, the dashboard CLI is **forgiving** — it always renders whatever data is available. Pre-flight is implicit, not blocking.
+
+---
+
+## Output structure
+
+```
+qa/dashboard/
+├── index.html                      # share-ready single-file report
+└── history/
+    ├── 2026-04-26T...Z.json        # past snapshots, oldest → newest
+    ├── 2026-04-27T...Z.json
+    └── …                           # max 20, older pruned automatically
+```
+
+Both should be committed to git so collaborators see the same trend lines.
+
+---
+
+## Dashboard views
+
+| View | What it shows |
+|------|---------------|
+| **Overview** | Stat cards (total/passed/failed/pending), pass-rate donut, priority bar chart, trend mini-chart, per-screen progress bars |
+| **Suites** | Tree: screen → scenarios; filter by search/status/priority; click → detail modal with steps, expected, error, trace |
+| **Trends** | Pass-rate line chart + stacked status bars across all retained runs (requires ≥2 runs) |
+| **Compare** | Pick base/head run; per-screen Δ table; lists of newly passing/failing/added/removed/changed tests |
+| **Export** | Browser-side CSV / XLSX download. One-button generation, no server. Same column layout as `sungen delivery` |
+
+Trends + Compare are disabled in the sidebar until at least 2 runs exist in history.
+
+---
+
+## CLI command reference
+
+```bash
+# All screens + flows
+sungen dashboard
+
+# Specific targets
+sungen dashboard kudos awards flow/checkout
+
+# Skip history persistence (one-off / CI ephemeral)
+sungen dashboard --no-history
+
+# Cap retained history (default: 20)
+sungen dashboard --max-history 50
+
+# Auto-open in default browser
+sungen dashboard --open
+```
+
+---
+
+## Skill responsibilities (when invoked from /sungen:dashboard)
+
+1. **Run the CLI** with whatever arguments came from `$ARGUMENTS`.
+2. **Show the CLI output verbatim** — do not summarize, paraphrase, or omit warnings.
+3. **Offer next-step options via `AskUserQuestion`** based on what just happened:
+   - On success → open dashboard, share file, run more tests, build for another screen, done.
+   - On error → diagnose from the error message; common causes are missing `qa/` directories or no targets.
+4. **Do NOT** parse `.feature`, `.spec.ts`, or results files yourself — that is the CLI's job.
+5. **Do NOT** touch files under `qa/dashboard/` directly — only the CLI writes there.
+
+---
+
+## Sharing the dashboard
+
+The HTML is self-contained:
+- No CDN or external font — system font stack only.
+- All JS, CSS, charts, and the SheetJS library are inlined.
+- The snapshot payload sits in `<script id="__SUNGEN_DASHBOARD__" type="application/json">…</script>`.
+
+Acceptable sharing: email, Slack upload, S3 bucket, committed in git. The recipient opens it directly, no install required. ~1MB on disk, ~300KB gzipped over the wire.

package/src/orchestrator/templates/playwright.config.ts

@@ -28,7 +28,14 @@ export default defineConfig({
   /* Output file path is controlled by PLAYWRIGHT_JSON_OUTPUT_NAME env var for per-screen isolation. */
   reporter: [
     ['html'],
-    ['json', { outputFile: process.env.PLAYWRIGHT_JSON_OUTPUT_NAME || 'test-results/results.json' }],
+    [
+      'json',
+      {
+        outputFile:
+          process.env.PLAYWRIGHT_JSON_OUTPUT_NAME ||
+          'test-results/results.json',
+      },
+    ],
   ],
   /* Shared settings for all the projects below. See https://playwright.dev/docs/api/class-testoptions. */
   use: {
@@ -40,6 +47,18 @@ export default defineConfig({
 
     /* Collect trace when retrying the failed test. See https://playwright.dev/docs/trace-viewer */
     trace: 'on-first-retry',
+
+    /**
+     * Capture screenshot after each test failure.
+     * See https://playwright.dev/docs/api/class-testoptions#test-options-screenshot
+     */
+    screenshot: 'only-on-failure',
+
+    /**
+     * Record video for each test, but remove all videos from successful test runs.
+     * See https://playwright.dev/docs/api/class-testoptions#test-options-video
+     */
+    // video: 'retain-on-failure',
   },
 
   /* Configure projects for major browsers */