request-refactor-plan
Pass
Audited by Gen Agent Trust Hub on Mar 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or unauthorized commands were detected. The skill provides a structured process for code exploration and documentation generation.
- [PROMPT_INJECTION]: The skill processes user input and codebase content to generate refactor plans. While this represents a surface for indirect prompt injection, it is inherent to the skill's primary purpose and the instructions include verification steps (e.g., exploring the repo to verify assertions) and user interviewing to refine the scope.
- Ingestion points: User description (Step 1) and codebase content (Step 2, Step 6).
- Boundary markers: None explicitly defined.
- Capability inventory: Repository exploration (read) and GitHub issue creation (write).
- Sanitization: Not explicitly mentioned, but the skill advises against including specific code snippets in the output issue, which mitigates some risks.
- [DATA_EXPOSURE_EXFILTRATION]: The skill reads codebase information and writes to GitHub. This is consistent with its stated purpose. No access to sensitive files (e.g., .env, .ssh) or exfiltration to unknown domains was found.
Audit Metadata