The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted codebase content and instructs the agent to execute code examples for verification during the self-verification phase. This creates a surface for indirect prompt injection where malicious content in the source files could be reflected into documentation and subsequently executed by the agent.
Ingestion points: Existing codebase files and project data analyzed during the documentation process.
Boundary markers: No specific delimiters or warnings are provided in the skill instructions to prevent the agent from obeying instructions embedded in the code being documented.
Capability inventory: The skill requires the agent to execute code examples and installation steps in a clean environment to verify documentation accuracy.
Sanitization: No sanitization, escaping, or validation of the source code content is described before it is included in documentation or executed.

doc-writer