The Agent Skills Directory

[SAFE]: The skill is restricted to generating planning documentation and directory structures within the 'plans/' folder.
[SAFE]: It implements a human-in-the-loop workflow, requiring user validation for requirements and final plan approval via the 'question' tool.
[SAFE]: Codebase analysis is performed through a specialized tool ('doc-explorer') with results stored in the 'docs/' directory, preventing direct execution of external code content.
[SAFE]: There are no instances of network requests, hardcoded credentials, or attempts to bypass agent safety guidelines.
[NO_CODE]: The skill consists exclusively of markdown instructions and templates without any executable scripts.

create-plan