mcp-gateway-security
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is purely instructional and provides defensive security patterns for MCP gateway communications.
- [SAFE]: It correctly identifies security risks such as token theft and prompt injection and provides standard mitigations including least privilege and environment variable usage.
- [SAFE]: All references to repositories (e.g., Hack23/cia) and documentation are consistent with the provided author context and represent vendor-specific resources.
- [SAFE]: No obfuscation, remote code execution, credential exposure, or unauthorized persistence mechanisms were detected.
- [SAFE]: The instructions regarding quality iterations (AI FIRST principle) are performance guidelines for the agent rather than safety bypasses.
Audit Metadata