| name | smoke-test |
| description | Post-sprint validation. Verifies the application works as expected after a sprint completes. Use after any sprint or deployment, or when the user says "smoke test", "verify it works", "test the app", "does it still work", "regression check", "post-deploy check", or "sanity check". Invoked by SP-QA during post-sprint review.
|
Smoke Test
You are running a post-sprint validation. Your job is to verify the
application still works correctly after recent changes.
Step 1 — Understand what changed
Read the latest sprint plan to understand what was modified:
tasks/sprints/sprint-*/plan.md (latest).
If tasks/sprints/sprint-*/build-output.md exists, read it for the
list of modified files.
Step 2 — Run the test suite
Detect the test runner:
- If
package.json exists → npm test or check scripts
- If
pytest.ini, setup.cfg, pyproject.toml → pytest
- If
Makefile with test target → make test
- If
Cargo.toml → cargo test
- If none found → report "no test runner detected"
Run the full test suite and capture results.
Step 3 — Compare with baseline
Check tasks/lessons.md or previous sprint outputs for baseline test counts.
If a baseline exists, compare:
- Are there fewer tests now? (tests removed?)
- Are there new failures?
Step 4 — Health checks
If the project has identifiable health endpoints or CLI commands:
- Web app: check if key endpoints respond (look for routes in code)
- CLI: verify key commands run without error
- API: check health/status endpoint if it exists
If none are identifiable, skip this step and note it.
Step 5 — Check for common regressions
- Imports that break (look for import errors in test output)
- Config files that changed (diff against previous sprint)
- Dependency conflicts (check lock file changes)
Step 6 — Produce output
Save to tasks/sprints/sprint-XX/smoke-test-output.md:
# Smoke Test — Sprint [XX]
## Test suite
- Runner: [pytest/npm/etc.]
- Total tests: [count]
- Passed: [count]
- Failed: [count]
- Baseline (last sprint): [count if known]
## Regression check
- New test failures: [list or "none"]
- Missing tests (removed?): [list or "none"]
## Health checks
- [Endpoint/command]: [status]
## Verdict
[PASS / FAIL / PASS WITH WARNINGS]
[If fail: what specifically failed and recommended action]
Channel mode (remote execution)
If invoked via Channel (Telegram/Discord), adapt your output:
- Send a SHORT summary (5-10 lines max) via the Channel reply
- Full output goes in the sprint file as usual
- Include the verdict prominently
- Keep Channel messages concise — the human is likely on a phone
Gotchas
- Don't skip the test suite. Running tests is the minimum bar.
- If tests take too long (>5 min), note the duration as a concern.
- A passing test suite doesn't mean the app works. Health checks catch what unit tests miss.
- If no test runner is detected, flag this as a major concern — the project has no automated tests.
STOP. Your deliverable is smoke-test-output.md.
Do NOT fix any issues found. Report them for the next sprint or /fix.
