The Agent Skills Directory

[COMMAND_EXECUTION]: The skill leverages the Bash tool to perform local file operations and automation. Specifically, it uses grep to verify agent IDs in agents/_registry.yml, cp to initialize new team files from templates, and npm run update-readmes to regenerate documentation via a local task runner.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface (Category 8) by design.
Ingestion points: User-provided inputs for 'Team purpose', 'role', and 'responsibilities' are collected in Steps 1, 3, and 6.
Boundary markers: The skill does not explicitly define delimiters (e.g., XML tags or triple quotes) to wrap user-provided prose when writing to the destination markdown files.
Capability inventory: The skill possesses Write, Edit, and Bash capabilities, allowing it to modify registry files and create executable agent configurations.
Sanitization: There is no evidence of input validation or escaping for user-provided text before it is interpolated into the team definition files.
[EXTERNAL_DOWNLOADS]: The procedure recommends running npm install to ensure the js-yaml dependency is available. While this involves downloading from the well-known NPM registry, it is a standard practice for maintaining the skill's supporting scripts.

create-team