The Agent Skills Directory

[SAFE]: The skill consists of pedagogical content designed to assist with modern Python development. All code snippets provided in SKILL.md, EXAMPLES.md, and REFERENCE.md follow standard security and development practices.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface typical of development assistants that process external data and code.
Ingestion points: User-supplied Python source code for refactoring and Pydantic model definitions for FastAPI request handling (referenced in EXAMPLES.md and REFERENCE.md).
Boundary markers: Absent; the skill does not define explicit delimiters or "ignore previous instructions" guards for the data it processes.
Capability inventory: The patterns include file system interactions via Repository patterns, network communication via asyncpg and httpx, and task queueing with Celery.
Sanitization: Present; the skill strongly emphasizes the use of Pydantic for strict input validation, type annotations for static analysis, and JWT for secure authentication, which serve as primary sanitization mechanisms.

python-pro