The Agent Skills Directory

[SAFE]: The skill provides structured guidance and templates for creating documentation. It adheres to best practices by emphasizing discovery-based content generation over assumptions.
[COMMAND_EXECUTION]: The skill utilizes standard, local discovery tools such as ripgrep (rg), find, and grep to analyze project structure. These operations are read-only and limited to project configuration files.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from a user's repository (e.g., package.json content) to generate documentation. While this creates a theoretical injection surface, the risk is minimized by the skill's instructions to focus on concrete paths and validated commands. There is no evidence of executable code being generated from these inputs beyond documentation strings.

agents-md