request-refactor-plan

Pass

Audited by Gen Agent Trust Hub on Jul 27, 2026

Risk Level: SAFEDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from the repository being analyzed and from user input to generate its refactoring plan.
  • Ingestion points: Repository exploration and code analysis (SKILL.md, steps 2 and 6).
  • Capabilities: Creating GitHub issues and reading local files.
  • Boundary markers: Uses a structured Markdown template for the final output.
  • Sanitization: Includes a directive to exclude code snippets and file paths from the final report.
  • [DATA_EXFILTRATION]: The skill is designed to transmit a generated refactor plan to an external service (GitHub).
  • The instructions provide a specific safeguard: "Do NOT include specific file paths or code snippets. They may end up being outdated very quickly." This instruction acts as a security boundary to prevent the exfiltration of sensitive source code or internal path structures.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 27, 2026, 10:24 AM
Security Audit — agent-trust-hub — request-refactor-plan