review

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • PROMPT_INJECTION (LOW): Indirect Prompt Injection surface detected. The skill processes external, untrusted code which could contain malicious instructions designed to influence the review report or the agent's subsequent actions.
  • Ingestion points: Processes user-provided file paths via $ARGUMENTS and the contents of files identified in the project scope (SKILL.md Phase 1).
  • Boundary markers: Absent. No specific boundary markers, delimiters, or 'ignore embedded instructions' warnings are used when passing file content to sub-skills.
  • Capability inventory: Orchestrates multiple review skills ($review-bugs, $review-type-safety, etc.) and presents a consolidated markdown report.
  • Sanitization: No sanitization of code content is performed before processing or reporting.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:20 PM