The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to audit user-provided Python code and diffs, creating an attack surface where instructions embedded in code comments or strings could influence the agent.\n
Ingestion points: Audits code, diffs, and repositories as specified in SKILL.md.\n
Boundary markers: No explicit instructions are provided to the agent to ignore or delimit instructions found within the audited code.\n
Capability inventory: The agent executes ruff and runs local Python audit scripts via pytest, which in turn invoke uv and pip-audit.\n
Sanitization: No sanitization is performed on the ingested code before it is processed by the agent or tools.\n- [COMMAND_EXECUTION]: The skill executes external tools to perform dependency and security audits.\n
Evidence: The scripts scripts/test_pypi_security_audit.py and scripts/test_uv_security_audit.py utilize subprocess.run to call pip-audit and uv audit. These calls use argument lists and avoid shell=True, following security best practices to prevent command injection from within the skill's own logic.

python