The Agent Skills Directory

[SAFE]: The skill documents defensive security audit patterns for Django settings and views, covering critical configurations such as SECRET_KEY, DEBUG, and CSRF protection.\n- [SAFE]: The provided bash script scripts/scan.sh facilitates localized static analysis of source code using ripgrep to detect potential vulnerabilities. The script is self-contained, does not perform network operations, and relies on standard shell utilities.\n- [INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it processes project files during auditing. \n
Ingestion points: The scripts/scan.sh utility reads content from a user-specified directory and project files (e.g., settings.py, views.py).\n
Boundary markers: The scanning script does not implement specific boundary markers as it is designed for localized string matching.\n
Capability inventory: The capabilities identified include file discovery via find and pattern matching via ripgrep; no network access, arbitrary execution, or file writing capabilities were found in the skill's scripts.\n
Sanitization: The scanning script properly quotes variables to prevent command injection during directory traversal and file reading.

security-django