code-reviewer

Pass

Audited by Gen Agent Trust Hub on Mar 24, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill requests access to the Read, Grep, and Glob tools to navigate and analyze the codebase. This access is necessary for its stated purpose of identifying bugs and vulnerabilities.
  • [PROMPT_INJECTION]: Instructions define the agent's role as a senior engineer and provide strict behavioral guidelines. These instructions are intended to ensure professional and thorough reviews and do not attempt to override safety guardrails.
  • [DATA_EXFILTRATION]: The skill processes local source code to provide feedback. There are no instructions or tools that facilitate the exfiltration of this data to external servers.
  • [SAFE]: The skill follows security best practices for AI agents, using local tools for analysis and providing references for common vulnerabilities. Note: While the skill ingests untrusted code (Ingestion points: PR descriptions and file content in SKILL.md; Boundary markers: Absent; Capability inventory: Read, Grep, Glob; Sanitization: Absent), its functionality is limited to analysis without execution, presenting minimal risk.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 24, 2026, 01:36 AM
Security Audit — agent-trust-hub — code-reviewer