Show HN: Cordon – Security gateway for MCP tool calls with HITL approvals

Category: security

Tags: security, mcp, ai-agents, gateway, typescript

Score: 7.3/10 (Innovation: 7, Technical: 7, Documentation: 7, Utility: 8)

Cordon is a security gateway for MCP (Model Context Protocol) that adds policy enforcement, audit logging, and human-in-the-loop approvals between AI agents and their tool servers. It solves a critical production blocker by providing granular access control for AI agent tool calls, including SQL-aware policies and hidden tool capabilities.

Target audience: backend devs, devops, security engineers

Repository: https://github.com/marras0914/cordon · TypeScript · MIT

View on Hacker News