The Agent Skills Directory

[SAFE]: The skill uses standard Python libraries such as ast, inspect, and glob to parse code structure and metadata for documentation purposes without executing the code itself.
[EXTERNAL_DOWNLOADS]: External resources referenced in the templates, such as jsDelivr for Swagger UI and official GitHub Actions (e.g., actions/checkout), target well-known and reputable services standard in software development workflows.
[DATA_EXPOSURE]: The skill explicitly instructs the agent to avoid exposing secrets, internal URLs, or sensitive data within generated documentation. Credential placeholders like YOUR_API_KEY and JWT_SECRET are used appropriately in examples.
[INDIRECT_PROMPT_INJECTION]: While the skill analyzes untrusted codebase content (ingestion point: APIDocExtractor.extract_endpoints in implementation-playbook.md), it uses static analysis via AST parsing rather than dynamic execution. Given its primary purpose of documentation, the risk is negligible.

documentation-generation-doc-generate