The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local Python script and Git-related commands to perform inventory and deployment tasks.
It runs scripts/inventory.py to aggregate skill metadata into a report.
It uses git and gh tools to create worktrees, commit updates, and open draft pull requests for skill improvements.
[INDIRECT_PROMPT_INJECTION]: The skill analyzes data from potentially untrusted local files, creating a surface for indirect instructions to influence the agent.
Ingestion points: The inventory.py script reads markdown and YAML content from all skill files in ~/.claude/ and the project's local configuration.
Boundary markers: No specific delimiters or safety instructions are provided to the agent to treat audited content as untrusted data.
Capability inventory: The skill can modify configuration files in the user's home directory and perform network operations via Git/GitHub CLI tools.
Sanitization: The skill does not sanitize or escape the harvested metadata before presenting it to the agent for auditing.

skill-doctor