The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes scripts init_skill.py and package_skill.py that perform administrative file system tasks. These operations include creating directory structures, writing template file content, and generating ZIP archives. These actions are necessary and expected for the skill's purpose as a project scaffolding and packaging utility.
[PRIVILEGE_ESCALATION]: The scripts/init_skill.py script applies executable permissions (chmod 0o755) to a dynamically created template script. While modifying permissions on generated files is often scrutinized, in this context it is a standard convenience for developer tooling to ensure created templates are immediately functional.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-defined strings for skill names and descriptions, which are then interpolated into documentation and scripts. This creates a surface for indirect prompt injection if untrusted skill metadata is processed.
Ingestion points: Metadata fields ingested from CLI arguments in init_skill.py and read from SKILL.md in quick_validate.py.
Boundary markers: The system uses standard YAML frontmatter delimiters (---) to isolate metadata.
Capability inventory: File system write access, directory creation, permission modification, and ZIP file generation.
Sanitization: quick_validate.py implements validation checks that restrict the use of angle brackets and enforce specific naming conventions (hyphen-case) for skill identifiers.

create-skill