---
name: run-checklist
description: Run a test checklist — execute items manually via browser and API, report results
---

Execute the test checklist generated by pr-analyst.

## Prerequisites (manual, before running this skill)

Before running, the tester must:
1. Check available stands: `/betas` in Slack (Bug Catcher bot)
2. Book a free stand (alpha / beta / gamma / delta / dashly)
3. Deploy the PR branch to that stand via Jenkins UI
4. Then run: `/run-checklist 5449 gamma`

## Input

Argument can be:
- PR number + stand: `/run-checklist 5449 gamma` — recommended
- PR number only: `/run-checklist 5449` — will ask which stand to use
- No argument: `/run-checklist` — use latest checklist, ask for stand
- File path + stand: `/run-checklist path/to/file.md gamma`

Stand URLs:
- alpha → https://alpha.k8s-dev.carrotquest.io
- beta → https://beta.k8s-dev.carrotquest.io
- gamma → https://gamma.k8s-dev.carrotquest.io
- delta → https://delta.k8s-dev.carrotquest.io
- dashly → https://dashly.k8s-dev.carrotquest.io

## Step 1 — Find checklist

```bash
# Latest file
ls -t ~/carrot-quest-qa/agents/pr-analyst/checklists-generated/*.md | head -1

# By PR number
ls ~/carrot-quest-qa/agents/pr-analyst/checklists-generated/ | grep "PR-5449"
```

Read the full checklist before starting.

## Step 2 — Execute checklist items

For each item — do what you can programmatically, test the rest via browser:

- **API / backend checks** — make direct curl requests, verify status codes and response body
- **UI checks** — describe exact steps for the human QA to verify manually; mark as PENDING
- **Side effects** — check via API that the expected state change happened

Mark each item: `PASS`, `FAIL`, or `SKIP` (with reason).

For any `FAIL` — immediately call `/create-bug-report`.

## Step 3 — Save report

Save full report to `test-results/YYYY-MM-DD_<CAR-id>_results.md`

Use template: `shared/templates/test-results.md`

## Step 4 — Print final summary

After saving the report, immediately output to the terminal:

```
✅ Done. Report saved: test-results/YYYY-MM-DD_<CAR-id>_results.md

Results: X passed / X failed / X skipped

  ✅ [item description]
  ❌ [item description] — bug filed
  ⏭️ [item description] — skipped: reason

Bugs found: X
```

Do NOT wait for the user to ask. Output this summary automatically when work is complete.