The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes pnpm install within a user-specified directory to initialize the documentation project. This is a standard operation for setting up the VitePress environment as described in the workflow.
[EXTERNAL_DOWNLOADS]: The skill downloads and installs Node.js packages including vitepress, mermaid, and vitepress-plugin-mermaid from the official NPM registry. These are well-known packages necessary for the skill's stated purpose.
[INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data by reading and analyzing local source code to generate descriptive documentation.
Ingestion points: The Task and Explore tools are used to read arbitrary files within the target project directory during the Deep Analysis phase (SKILL.md).
Boundary markers: The skill does not define specific delimiters or instructions to prevent the agent from following directions potentially embedded in comments or strings within the analyzed source code.
Capability inventory: The skill has the ability to write files (Markdown documentation, configuration files), create directory structures, and execute shell commands (pnpm install).
Sanitization: There is no evidence of sanitization or filtering applied to the content extracted from source files before it is used to generate the VitePress site.

vitepress-tutorial