The Agent Skills Directory

[NO_CODE]: The skill consists exclusively of markdown documentation and a topic index.
It does not include any standalone executable scripts, binaries, or background services.
All content is static reference material.
[COMMAND_EXECUTION]: The skill instructs the agent to use standard CLI utilities to interact with the documentation.
Evidence in SKILL.md directs the agent to use grep and cat to search the references/ directory.
This behavior is restricted to the skill's own content and is necessary for the stated purpose of answering documentation questions.
[PROMPT_INJECTION]: The skill includes instructions to ensure the agent stays grounded in the documentation.
SKILL.md contains a 'CRITICAL' marker instructing the agent to 'grep the references/ directory' and 'Do NOT guess'.
This is a benign grounding pattern used to reduce model hallucination and ensure accurate technical support.
[EXTERNAL_DOWNLOADS]: The documentation contains references to official installation commands and trusted repositories.
Files such as overview.md and setup.md reference official scripts from claude.ai and anthropic.com.
Per standard operating procedures for documentation skills, these links to well-known services are considered informative and safe.

claude-code-docs