Test
Comprehensive testing framework for code and UI. Runs the full test stack, analyzes failures, measures coverage, and produces QA reports.
Core Principle
NEVER IGNORE FAILING TESTS.
No workarounds. No mocks to pass CI. No --passWithNoTests. Fix the root cause or document a legitimate skip with justification.
What This Skill Does
Test orchestrates the complete quality verification loop: type checking, test execution, failure analysis, coverage measurement, UI validation, and report generation. Works with any test runner and integrates with the debug skill for failures requiring investigation.
Workflows
Code Testing
Supports all major test runners:
| Language | Runners |
|---|---|
| JavaScript/TypeScript | Jest, Vitest, Mocha |
| Python | pytest, unittest |
| Go | go test |
| Rust | cargo test |
| Dart/Flutter | flutter test |
UI Testing
Delegates to the chrome-devtools skill for:
- Visual regression
- Interaction testing
- Accessibility checks
- Performance audits
Report Generation
Produces structured QA reports at {active-plan}/reports/tester-YYMMDD.md covering: pass/fail summary, coverage metrics, failing test details, and recommendations.
Working Process
Identify Scope → Typecheck First → Execute Suites
→ Analyze Failures → Coverage Report → UI Tests (if needed)
→ QA Report- Identify Scope — determine which suites are relevant (unit/integration/e2e/UI)
- Typecheck First — run
tsc --noEmitor equivalent before tests - Execute Suites — run test runner(s) for the project
- Analyze Failures — for each failure: read error, trace root cause, recommend fix
- Coverage Report — Istanbul/c8/pytest-cov/go cover output
- UI Tests — browser-based validation via chrome-devtools skill
- QA Report — write structured report with actionable findings
Tools Integration
| Tool | Purpose |
|---|---|
| Test runners | Execute suites (auto-detected from package.json / config files) |
| Istanbul/c8 | JS/TS coverage |
| pytest-cov | Python coverage |
| go cover | Go coverage |
| chrome-devtools skill | Browser/UI testing |
| ai-multimodal skill | Screenshot analysis for UI failures |
| debug skill | Root cause analysis for complex failures |
Quality Standards
- Critical paths: 100% test coverage required
- Happy + error paths: both must be covered
- Test isolation: no shared state between tests
- Deterministic: tests must produce same result every run
- Data cleanup: tests clean up any created data
- No flaky tests: fix or remove, never retry-loop
Team Mode
In agent team sessions, the tester teammate:
- Owns all test files exclusively
- Reads implementation files but never edits them
- Reports coverage and failures back to lead
Example Prompts
- “Run tests for the auth module and generate a coverage report”
- “Run the full test suite and fix any failures”
- “Run UI tests on the dashboard page and check for regressions”
- “Run integration tests and analyze why the payment tests are failing”
Related Skills
- Debug — activated for failures requiring root cause analysis
- Cook — cook’s quality gate requires 100% test pass
- Chrome DevTools — UI and browser testing