audit-idiomatic
Pass
Audited by Gen Agent Trust Hub on Mar 10, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: No malicious patterns or security risks detected in the skill's instructions.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted codebase data (source code and project manifests like package.json, pyproject.toml, etc.) and possesses file-modification capabilities through its auto-fix feature. Ingestion points: Repository source code and configuration manifests. Boundary markers: Absent. Capability inventory: File-write (auto-fix). Sanitization: Relies on LLM instruction to ensure changes are behavior-preserving and requires human approval for architectural shifts.
Audit Metadata