guided-contribution
Pass
Audited by Gen Agent Trust Hub on Jul 10, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is entirely instructional and does not include executable code or risky configurations.
- [SAFE]: No patterns of prompt injection, data exfiltration, or privilege escalation were found. The skill encourages user-led verification of all codebase changes.
- [SAFE]: While the skill interacts with untrusted codebase data, its requirement for manual review and multi-level behavior verification (Phase 4) serves as a robust security control against indirect prompt injection.
Audit Metadata