skills/melumuccu/ai/genshijin-review/Gen Agent Trust Hub

genshijin-review

Pass

Audited by Gen Agent Trust Hub on Jul 6, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill contains no executable code or malicious logic, serving only as a set of formatting guidelines for pull request reviews.
  • [PROMPT_INJECTION]: The skill defines specific activation triggers and formatting constraints but does not attempt to override the underlying model's safety or ethical guardrails.
  • [DATA_EXFILTRATION]: No network operations, credential harvesting, or unauthorized file access patterns were identified.
  • [REMOTE_CODE_EXECUTION]: The skill does not install packages or execute remote scripts.
  • [PROMPT_INJECTION]: The skill processes untrusted code (PR diffs) as its input; however, the risk is negligible as it is strictly limited to generating text-based review comments and lacks the capability to execute commands or modify files.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 6, 2026, 08:04 AM
Security Audit — agent-trust-hub — genshijin-review