council
Pass
Audited by Gen Agent Trust Hub on Jun 27, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill functions as a facilitator for local subagents, using the Agent tool to dispatch archetypes defined within the project structure (e.g.,
.claude/agents/). This does not involve remote code execution or untrusted package installation. - [SAFE]: No network operations (curl, wget, etc.) or sensitive file access patterns (credentials, SSH keys) were detected. The skill's operations are confined to processing user input and managing local agent interactions.
- [SAFE]: The skill includes internal 'Debate Protocols' that act as a quality and safety control mechanism. It monitors subagent responses for 'authenticity' and 'protocol violations,' with instructions to re-dispatch if an agent deviates from its assigned role or safety constraints.
- [SAFE]: No obfuscation techniques such as Base64, hex encoding, or zero-width characters were found in the instructions or reference files.
Audit Metadata