The Agent Skills Directory

[DYNAMIC_EXECUTION]: The scripts/init_skill.py script generates executable Python files from internal templates, populating them with user-supplied skill names.
[PRIVILEGE_ESCALATION]: The scripts/init_skill.py script programmatically makes generated scripts executable by calling chmod(0o755).
[COMMAND_EXECUTION]: The skill provides instructions for the agent to execute local scripts (init_skill.py and package_skill.py) to manage the skill lifecycle.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied identifiers and descriptions to generate system-level instruction files, creating an injection surface.
Ingestion points: User-provided <skill-name> and <path> arguments in scripts/init_skill.py.
Boundary markers: The generated files use YAML frontmatter and Markdown headers to delimit user-provided content.
Capability inventory: The skill can create directories, write files, and modify file permissions on the local filesystem.
Sanitization: The scripts/quick_validate.py script enforces hyphen-case naming conventions and filters angle brackets from metadata fields to prevent basic injection attacks.

skill-creator