django-verification

Run a phased verification covering environment checks, linting with ruff and black, migration validation with manage.py makemigrations --check, pytest with coverage, and security scans via pip-audit and bandit. Finish with manage.py check --deploy to confirm production settings.

Use pytest with the pytest-cov plugin: pytest --cov=apps --cov-report=html --cov-report=term-missing. This generates terminal and HTML reports showing per-app coverage, with typical targets of 80% overall and 90% for models and services.

pip-audit and safety check scan installed Python packages against known vulnerability databases, while bandit performs static analysis of Django source code for common security issues. gitleaks can additionally detect hardcoded secrets in the repository.

Yes, the skill includes a GitHub Actions workflow that sets up Python, caches pip dependencies, runs ruff, black, isort, and mypy checks, executes bandit, safety, and pip-audit scans, and runs pytest with coverage against a PostgreSQL service container.

Failures usually come from model changes without generated migrations, unapplied migrations, or conflicting migration branches. Use manage.py makemigrations --check to detect missing migrations and manage.py makemigrations --merge to resolve conflicts.

Automated checks catch common issues like lint violations, low coverage, and vulnerable dependencies, but they do not replace manual code review or staging environment testing. Business logic errors and integration issues still require human validation.