| name | run-all-tests |
| description | Use when the user asks to 'run all tests', 'test everything', 'run the full pipeline', 'check all stages', or 'run test-all'. Runs scripts/test-all.sh and reports per-tier results with a summary table. |
| disable-model-invocation | true |
Run All Tests — Full Pipeline
Runs the complete Buddy Evolver test pipeline (all tiers in test-all.sh) and
reports per-tier results with pass/fail counts, durations, and failure details.
Steps
1. Run test-all.sh
cd "${CLAUDE_PLUGIN_ROOT}" && bash scripts/test-all.sh 2>&1
2. Read results.json
cat "${CLAUDE_PLUGIN_ROOT}/test-results/results.json"
3. Display summary table
Format one row per tier from results.json. Use ✅ if exit_code == 0,
❌ otherwise. Group by stage field when present.
Test Pipeline Results
══════════════════════════════════════════════════════════
TIER STATUS PASSED DURATION
──────────────── ────── ──────── ─────────
smoke ✅ 13/13 8s
unit ✅ 98/98 12s
security ✅ 25/25 3s
ui ✅ 23/23 2s
snapshots ✅ 6/6 1s
docs ✅ 16/16 1s
──────────────────────────────────────────────────────
TOTAL ✅ 181/181 27s
4. On failure
If any tier has exit_code != 0, show which tier failed and suggest:
- smoke failure → build or CLI contract broken; check
swift build output
- unit failure → run
/run-tests to see per-suite breakdown with failure details
- security failure →
bash scripts/test-security.sh — run standalone for focused output
- snapshots failure → if CLI output changed intentionally, run
UPDATE_GOLDEN=1 bash scripts/test-snapshots.sh to regenerate golden files; review diffs before committing
- docs failure → run
/sync-docs to fix documentation drift