agentic-review-handoff

Pass

Audited by Gen Agent Trust Hub on May 15, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is designed specifically for read-only code review tasks. It contains explicit instructions for the agent to avoid editing files, committing, or pushing code unless specifically switched to a 'fix mode' by the user.
  • [PROMPT_INJECTION]: The 'Feedback Validation' mode includes defensive instructions to treat external reviewer feedback as unverified claims. The agent is directed to verify all claims against local source code and documentation, which serves as an effective mitigation against indirect prompt injection embedded in external review comments.
  • [DATA_EXFILTRATION]: No network operations or data exfiltration patterns were detected. The skill uses standard local commands such as git status and pwd purely for maintaining context and verifying the scope of the review.
  • [REMOTE_CODE_EXECUTION]: The skill does not perform remote downloads or execute untrusted scripts. It relies on the user's existing environment and test suites for verification, as documented in the handoff and contract references.
Audit Metadata
Risk Level
SAFE
Analyzed
May 15, 2026, 07:13 AM
Security Audit — agent-trust-hub — agentic-review-handoff