The Agent Skills Directory

[PROMPT_INJECTION]: Indirect prompt injection surface identified. The skill reads untrusted codebase content which could contain malicious instructions designed to influence the agent's behavior or poison the generated documentation. \n- Ingestion points: The skill reads package.json, README.md, project source code, and configuration files via Read, Glob, and Grep tools. \n- Boundary markers: There are no explicit delimiters or instructions provided to the agent to distinguish project data from its core instructions. \n- Capability inventory: The skill utilizes the Write tool to populate documentation in .chalk/docs/ and modify the AGENTS.md file. \n- Sanitization: No input validation or sanitization is performed on the content extracted from the codebase before it is incorporated into the documentation stubs.

setup-docs