The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill generates unit tests based on code and requirements provided in the user context, which is an inherent surface for indirect prompt injection.
Ingestion points: Python code snippets and function definitions supplied by the user in the agent prompt.
Boundary markers: The skill does not explicitly define markers or instructions to isolate user-provided code from the test generation instructions.
Capability inventory: The generated test code facilitates file system operations (via the tmp_path fixture), database transactions, and network mocking.
Sanitization: No input validation or sanitization is performed on the code provided for test generation.
[EXTERNAL_DOWNLOADS]: The CI/CD implementation guide references official and well-known GitHub Actions for automated testing workflows.
Evidence: The GitHub Actions configuration in reference/playbook.md references actions/checkout@v4, actions/setup-python@v5, and codecov/codecov-action@v4.

pytest-skill