coding-research
Pass
Audited by Gen Agent Trust Hub on Jul 24, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill instructions define a strict research-only boundary that prohibits the agent from creating, editing, or deleting files, effectively preventing unauthorized repository modifications.
- [SAFE]: Local inspection is limited to read-only tools like repository manifest analysis and code searches, with no evidence of data exfiltration or unauthorized network communication.
- [SAFE]: No obfuscated content, hardcoded credentials, or remote code execution patterns were detected in the skill markdown or agent configuration.
Audit Metadata