The Agent Skills Directory

[SAFE]: The skill's operations are transparent and strictly confined to the local project environment. It reads project metadata to build situational awareness and writes completed documentation to a local 'docs' directory, with no evidence of data exfiltration or unauthorized access.
[NO_CODE]: This is a no-code skill consisting entirely of instructional markdown and reference templates. It does not include any Python or Node.js scripts, binary executables, or external package dependencies.
[PROMPT_INJECTION]: The skill includes comprehensive 'Anti-Shortcut Rules' designed to prevent the AI from generating low-quality or fabricated content, which serves as a defensive measure against common model failure modes.

prd-generation