transplant-commits

Pass

Audited by Gen Agent Trust Hub on Jul 23, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill defines a structured workflow for git history management using standard local tools like cherry-pick and am.
  • [SAFE]: The instructions proactively mitigate indirect prompt injection risks by requiring the agent to treat repository metadata, commit messages, and mailbox content as untrusted data requiring independent verification.
  • [SAFE]: No indicators of data exfiltration, sensitive file access, or remote code execution were found in the instructions or references.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 23, 2026, 03:46 AM
Security Audit — agent-trust-hub — transplant-commits