review-validation

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • Indirect Prompt Injection Surface: The skill processes untrusted data from pull request metadata, descriptions, and repository files, which could potentially contain malicious instructions intended to influence the validation process.
  • Ingestion Points: Data enters the agent context through args.prBody, args.prTemplate, and repository content retrieved via the read_repo_file tool.
  • Boundary Markers: The skill includes a 'Security' section that explicitly instructs the agent to treat pull request content as untrusted and to disregard any instructions embedded within it.
  • Capability Inventory: The skill is equipped with tools to read repository files (read_repo_file), search the codebase (search_repo), and submit validation results (submit_review_validation).
  • Sanitization: The skill relies on natural language instructions to maintain boundaries rather than programmatic sanitization of the input data.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 04:04 PM
Security Audit — agent-trust-hub — review-validation