The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill documentation lists numerous standard Python packages for the MkDocs ecosystem, including mkdocs, mkdocs-material, and mkdocstrings. These are retrieved from PyPI, a well-known package registry, and are essential for the skill's stated functionality.
[REMOTE_CODE_EXECUTION]: Includes a reference to the official D2 diagramming tool installation script via curl -fsSL https://d2lang.com/install.sh | sh. While this pattern involves remote execution, the source is a recognized and established service for the specific task of diagram generation.
[DYNAMIC_EXECUTION]: Provides examples of Python scripts (gen_ref_pages.py) and hooks (my_hooks.py) used to automate the generation of documentation files and extend MkDocs' build process. These scripts are transparent, well-documented examples of standard developer workflows.
[COMMAND_EXECUTION]: The skill uses the Bash tool to execute legitimate MkDocs commands such as mkdocs build, mkdocs serve, and mkdocs gh-deploy. These operations are performed within the expected scope of a documentation management tool.
[INDIRECT_PROMPT_INJECTION]: By integrating tools like mkdocstrings that parse source code comments, the skill introduces a surface for processing untrusted external data. However, this is a core requirement of its primary function (API documentation generation), and no specific exploitable patterns were identified.

mkdocs