The Agent Skills Directory

[SAFE]: The skill serves as a development toolkit for generating, validating, and packaging agent skills. The included Python scripts perform standard local file system tasks such as directory creation and zip archive generation without network access or unauthorized data exposure.
[DYNAMIC_EXECUTION]: The init_skill.py script generates basic Python helper scripts from static internal templates. This script generation is a core, documented feature of the initialization process and does not involve executing untrusted or remote code.
[INDIRECT_PROMPT_INJECTION]: The skill presents a surface for indirect injection by interpolating user-provided inputs into template files. Evidence chain: (1) Ingestion points: skill_name and path arguments in init_skill.py. (2) Boundary markers: Absent. (3) Capability inventory: File system write operations in init_skill.py. (4) Sanitization: Basic validation in quick_validate.py, including checks for angle brackets and character length. The risk is considered safe given the intended purpose as a manual creation tool.

skill-creator