Security & trust layer for AI agents: scans an MCP server or skill before install for exfiltration, prompt-injection sinks, dangerous capabilities and obfuscation; also endpoint safety verdicts, vuln scans, threat intel and compliance checks (SOC 2/ISO 27001/HIPAA).
No one-line install is published for this server yet - see the repository for setup instructions.
https://wingmanprotocol.com/mcp.