Agentmesh - AI agent governance middleware

Name: Agentmesh - AI agent governance middleware
Author: angelnicolasc

@angelnicolasc

About Agentmesh - AI agent governance middleware

Governance middleware for AI agents: deterministic policy enforcement, cryptographic audit trails with digital signatures, DLP/PII detection, Trust Score per agent (0-100), EU AI Act compliance (Art. 9, 11, 12, 14), Agent BOM generation, and Circuit Breaker. Native support for La

#governance #compliance #security #audit #dlp #eu-ai-act #trust-score #policy-enforcement #langgraph #crewai #autogen #mcp

Basic information

Config

Add this server to your MCP-compatible client using the configuration below.

{
  "mcpServers": {
    "agentmesh": {
      "command": "uvx",
      "args": [
        "useagentmesh"
      ],
      "env": {
        "AGENTMESH_API_KEY": "your-api-key"
      }
    }
  }
}

Tools

No tools detected

We auto-extract tools from the README. The maintainer can list them under a ## Tools heading to populate this section.

Overview

What is Agentmesh - AI agent governance middleware?

Agentmesh is a governance platform for AI agents with two layers: a free offline CLI that scans codebases via AST to generate a governance score, Agent BOM, and EU AI Act gap analysis, and a paid SaaS runtime platform that enforces policies, scans payloads for PII, and tracks agent trust in production. It is designed for developers and teams building AI agents who need compliance, security, and observability.

How to use Agentmesh - AI agent governance middleware?

Install the CLI with pip install useagentmesh, then run agentmesh scan . for a governance report. To use the MCP server with Claude Desktop or VS Code/Cursor, add the provided JSON configuration to the appropriate settings file, setting the AGENTMESH_API_KEY environment variable for Claude Desktop.

Key features of Agentmesh - AI agent governance middleware

AST-based governance scoring with 33 deterministic policy rules (<2ms evaluation)
Agent BOM (bill of materials) for agents, tools, and models
Native SARIF 2.1.0 output for GitHub Code Scanning integration
EU AI Act gap detection (Art. 9, 11, 12, 14)
Runtime DLP with Presidio-based PII/PCI scanning
Dynamic trust score and circuit breaker for production agents

Use cases of Agentmesh - AI agent governance middleware

Scan a Python codebase using LangGraph or CrewAI to surface governance gaps before deployment
Enforce data loss prevention policies on tool call payloads in production
Generate EU AI Act compliance reports for regulatory audits
Suspend misbehaving agents automatically via trust score thresholds
Track and manage non-human identities across agent teams

FAQ from Agentmesh - AI agent governance middleware

What does the free CLI include vs the paid platform?

The free CLI offers offline scanning, governance score, Agent BOM, SARIF output, and EU AI Act gap detection. The paid SaaS platform adds runtime policy enforcement, DLP scanning, dynamic trust score, circuit breaker, ODD, magnitude limits, and agent identity management.

What are the runtime requirements?

The CLI requires Python 3.10+. The runtime platform requires an account and an API key for MCP integration.

Where does data live when using the CLI?

The CLI runs entirely offline with no account needed; no data is sent externally. The runtime platform is SaaS and processes data in the cloud.

How does agentmesh compare to alternatives like Bifrost or Cordum?

Agentmesh is Python-based and provides static governance scoring, Agent BOM, SARIF, and EU AI Act gap detection, which alternatives built in Go do not offer. The platform also includes runtime features missing from those tools.