Show HN: Imara – policy enforcement layer for MCP agents (npx imara)

Category: security

Tags: mcp, policy-enforcement, ai-agents

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

Imara is a policy enforcement layer for MCP agents that intercepts tool calls and applies YAML-defined rules to allow, deny, rate-limit, or escalate actions. It provides a hash-chained audit trail, a real-time dashboard, and zero-config integration, making it relevant for compliance with emerging AI regulations like the EU AI Act.

Target audience: backend devs, devops, data engineers

Repository: https://github.com/Dnakitare/imara · TypeScript · Apache-2.0

View on Hacker News