The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill provides templates for creating web applications that ingest and process untrusted data from external users via API endpoints.
Ingestion points: Data enters the system through Pydantic schemas in endpoints defined in references/details.md (e.g., create_user, update_user).
Boundary markers: The templates rely on standard Pydantic validation but do not include specific delimiters or instructions for the agent to ignore potentially malicious content embedded in user-supplied strings.
Capability inventory: The skill facilitates database write operations via SQLAlchemy and handles authentication logic.
Sanitization: Standard type validation and schema enforcement are provided by Pydantic; SQLAlchemy is used for database interactions, providing protection against SQL injection. The provided CORS middleware configuration in references/details.md uses a wildcard allow_origins=["*"], which is a permissive setting that developers should restrict in production.

fastapi-templates