| name | test-project |
| description | Execute comprehensive test suite with coverage reporting |
| license | MIT |
| compatibility | opencode |
| metadata | {"audience":"developers","workflow":"testing"} |
What I do
After implementing features or fixing bugs, I run the complete test suite:
- Unit tests - All pytest tests in the project
- Coverage analysis - Verify 80%+ code coverage
- Coverage reports - Generate HTML and XML reports
- Test parallelization - Run tests efficiently across CPU cores
- Failure reporting - Detailed output for debugging failures
When to use me
Use this skill after implementing features or fixing bugs:
- After writing new code features
- After refactoring components
- After fixing bugs
- Before submitting pull requests
- To verify code coverage thresholds
How I work
I execute: make test
This Makefile target runs pytest with coverage:
make test
Which runs:
uv run python -m pytest --cov --cov-config=pyproject.toml --cov-report=xml tests
Coverage requirements:
- Minimum 80% code coverage (
fail_under=80 in pyproject.toml)
- HTML reports generated to
coverage_html_report/
- XML reports for CI/CD integration
What I report
I'll show you:
- ✅ All tests pass - With coverage percentage
- ❌ Test failures - Which tests failed and why
- 📊 Coverage summary - Overall and per-module coverage
- 🔴 Coverage gaps - Which lines/functions lack tests
What to do if tests fail
If make test fails:
- Test errors: Review the pytest output for assertion failures
- Coverage gaps: Add tests for uncovered code paths
- Import errors: Check for missing test dependencies
- Fixture issues: Verify pytest fixtures are properly defined
Coverage best practices
- Aim for 80%+ coverage (enforced by config)
- Critical paths should have 100% coverage
- Use
@pytest.mark.coverage for critical tests
- Test edge cases, not just happy paths
Integration tips
After you implement code, ask me to:
- "Load the test-project skill and run tests"
- "Check if my new feature has adequate test coverage"
- "Run the full test suite on my changes"
Or I may automatically suggest this skill after you've written tests.
Related Skills: lint-project, check-project