The Agent Skills Directory

[SAFE]: The skill serves as a security auditing tool specifically designed for FastAPI developers. It provides a structured procedure for authorized 'self-pentests' to identify weaknesses like exposed OpenAPI schemas, missing security dependencies, and improper Pydantic field handling.
[SAFE]: While it describes 'attack' techniques such as JWT algorithm confusion (alg: none) and path traversal (..%2F..%2F.env), these are framed within a defensive auditing context (e.g., 'Authorized self-pentest probe').
[SAFE]: The skill identifies sensitive local files (like .env and pyproject.toml) as probe targets, but it does so to verify if a user's FileResponse routes are vulnerable to traversal, rather than attempting to exfiltrate them to a third party.
[SAFE]: All references lead to official documentation (Tiangolo, PyJWT) and well-known industry standards. No external downloads, remote scripts, or malicious code are present in the skill.

fastapi-attack-probe