skill-creator

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides functional Python scripts (scripts/init_skill.py and scripts/package_skill.py) that facilitate the creation of skill directories, writing of template files, and packaging into distributable formats. These scripts are intended for local execution by the user during the development workflow.
  • [INDIRECT_PROMPT_INJECTION]: The init_skill.py script serves as a template engine that interpolates a user-provided skill name into documentation and script templates. This represents an indirect injection surface where user-supplied strings are embedded into instruction files that may later be processed by an AI agent.
  • Ingestion points: Command-line arguments for and in scripts/init_skill.py.
  • Boundary markers: No specific delimiters are used to wrap the interpolated skill name within the generated templates.
  • Capability inventory: File system access (mkdir, write_text, chmod), and archive creation (ZipFile).
  • Sanitization: The script uses pathlib for path resolution and normalization, but does not explicitly sanitize the skill name string beyond basic formatting.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 09:17 AM
Security Audit — agent-trust-hub — skill-creator