testing-workflow

name	testing-workflow
description	Use when verifying CI, tests, and training integration. Runs the full test and lint suite and reports results.

Skill: testing-workflow

This skill verifies the repository's CI pipeline, tests, and training integration.

When to Use

• Before and after any code change
• When debugging failing CI
• To confirm the repository is in a healthy state
• Before opening a pull request

Commands

Run in this exact order:

# 1. Lint check
ruff check .
flake8 . --max-line-length=88 --extend-ignore=E203,W503

# 2. Format check
black --check .

# 3. Type check
mypy . --ignore-missing-imports

# 4. Unit tests
pytest tests/ -v --tb=short

# 5. Module import sanity check
python -c "import sys; sys.path.insert(0, 'src'); from model import cats_model; m = cats_model(pretrained=False); print('Model OK')"

Alternatively, use the verify script:

bash .agents/skills/testing-workflow/verify.sh

Expected Output

• All ruff check and flake8 commands exit with code 0
• All pytest tests pass (exit code 0)
• No import errors

Failure Handling

• If lint fails: fix code style issues before running tests
• If tests fail: read the traceback carefully, fix the source code, then re-run
• Never skip failing tests - fix them first

Integration with CI

Same commands run automatically in .github/workflows/train.yml on every push and PR.