The Agent Skills Directory

[EXTERNAL_DOWNLOADS] (SAFE): The documentation mentions standard library updates (e.g., Starlette, FastAPI) to address specific CVEs. These are presented as remediation advice rather than automated or risky downloads.
[COMMAND_EXECUTION] (SAFE): Includes an educational example of how to safely use the subprocess module by validating inputs with .isalnum() and using argument lists to prevent shell injection. No actual commands are executed by the skill's infrastructure.
[CREDENTIALS_UNSAFE] (SAFE): The code snippets use best practices such as os.environ for database URLs and standard placeholders for testing fixtures. No hardcoded secrets or sensitive credentials were found.
[PROMPT_INJECTION] (SAFE): No instructions targeting agent behavior override or safety bypass were detected. The content is strictly technical reference.
[DATA_EXFILTRATION] (SAFE): No network request patterns or sensitive file access paths consistent with data theft were identified.

fastapi