skill-creator
Pass
Audited by Gen Agent Trust Hub on Mar 8, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill invokes local Python scripts (audit.py, package.py, progress.py) using the Bash tool to perform scoring, packaging, and session management. These operations are confined to the local filesystem and the skill's own directory.
- [SAFE]: Analysis of the Python scripts confirms the use of safe library functions (e.g., yaml.safe_load) and local-only operations. No hardcoded credentials, sensitive file access (like .ssh or .env), or unauthorized network communications were found.
- [PROMPT_INJECTION]: The skill acts as an auditor for other skills, which involves reading external SKILL.md files. It mitigates indirect prompt injection risks by using regex-based deterministic analysis and providing a structured review protocol rather than executing instructions from audited content.
- [DYNAMIC_EXECUTION]: The progress.py script dynamically loads its companion audit.py script using importlib. This is used for internal modularity and is limited to loading a trusted file bundled within the same skill package.
Audit Metadata