The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute multiple local Python scripts for repository management and skill lifecycle tasks, including scripts/init_skill.py, scripts/quick_validate.py, scripts/sweep_skills.py, scripts/check_changed_files.py, and scripts/package_skill.py.
[INDIRECT_PROMPT_INJECTION]: The skill operates by transforming user-supplied task descriptions into instruction sets, creating an attack surface where malicious user input could inject harmful behavior into the generated skill's SKILL.md.
Ingestion points: User prompts and requests for initializing or editing skills.
Boundary markers: The documentation emphasizes clarity and brevity but does not mandate the use of boundary markers or instructions to ignore embedded commands in the generated content.
Capability inventory: Involves file creation and script execution via the scripts/ directory.
Sanitization: No sanitization or validation of user-provided intent is mentioned.
[DYNAMIC_EXECUTION]: The skill's workflow includes scaffolding new directories and potentially generating or modifying executable scripts (e.g., scripts/extract.py) as part of the skill development process.
[EXTERNAL_DOWNLOADS]: The skill notes a dependency on the PyYAML Python package, which is required for frontmatter validation tasks.

skill-creator