- Bernstein @chernistry/bernsteinDeclarative agent orchestration for engineering teams4% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from PyPI.1,416 GitHub stars.
- Bernstein @sipyourdrink-ltd/bernsteinThe open-source governance layer for AI agents. Byte-identical run receipts, 40+ adapters, air-gap.7% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from PyPI.1,215 GitHub stars.
- Agent Swarm @desplega-ai/agent-swarmFull Agent Swarm API and MCP server with local SQLite storage for multi-agent orchestration.6% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.873 GitHub stars.
- SSH — policy-gated remote access @tufantunc/ssh-mcpPolicy-gated, audited SSH for Linux and Windows hosts: roles, approvals, and an audit log.5% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.795 GitHub stars.
- HOL Guard @hashgraph-online/hol-guardLocal-first AI agent security evidence and approval workflows through HOL Guard's stdio MCP server.5% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from PyPI.785 GitHub stars.
- SandBase Harness @sandbaseai/sandbase-harnessStdio MCP bridge for SandBase Harness agents, sessions, turns, artifacts, and cancellation.5% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from Docker.707 GitHub stars.
- Decionis Commerce Gate MCP @decionis/commerce-gateCommerce preflights, Walmart mappings, D365 authorization, and offer checks; no external writes.7% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.687 GitHub stars.
- Vestige @samvallad33/vestigeLocal-first memory for AI agents that reaches backward to find a failure's root cause.5% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.649 GitHub stars.
- Emilia Protocol @emiliaprotocol/mcp-serverExact-action approval for consequential agent actions: request, track, and verify signed receipts.5% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.612 GitHub stars.
- Foremerge @naw103/foremergeCoordinate parallel coding agents, above Git5% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from cargo.538 GitHub stars.
- Vibe Check @pv-bhat/vibe-check-mcp-serverMetacognitive AI agent oversight: adaptive CPI interrupts for alignment, reflection and safety7% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.502 GitHub stars.
- Claudexor @razzant/claudexorRoute coding work across Claude Code, Codex, Cursor, and OpenCode with shared context.6% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.501 GitHub stars.
- Entroly @juyterman1000/entrolyContext Assurance for AI agents: context engineering, AI cost optimization, verification.6% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from PyPI.474 GitHub stars.
- Martin Loop @keesan12/martin-loopOne system around coding agents to run, verify, recover, and hand off software work.6% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.409 GitHub stars.
- MCPProxy @smart-mcp-proxy/mcpproxy-goLocal-first MCP proxy with BM25 tool discovery, security scanning, quarantine & ~99% token savings6% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from a package registry.387 GitHub stars.
- DashClaw @ucsandman/dashclawPolicy checks, approvals, records, and governed HTTP capabilities for unattended agents.6% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.312 GitHub stars.
- OpenWorkProof @dengyier/openworkproofVerifiable execution for AI agents: Ed25519-signed work contracts and audit trails.5% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from PyPI.288 GitHub stars.
- Sage @l33tdawg/sagePersistent, consensus-validated institutional memory for AI agents. Runs locally.9% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from Docker.255 GitHub stars.
- Codex Bridge @fanch-hui/codex-bridgeConnect MCP clients to local Codex Bridge agents, projects, approvals and workspace tools.11% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from MCPB.252 GitHub stars.
- Codex Bridge @yeyuancc0-glitch/codex-bridgeConnect MCP clients to local Codex Bridge agents, projects, approvals and workspace tools.8% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from MCPB.249 GitHub stars.
- Superlocalmemory @varun369/superlocalmemoryLocal-first AI memory with knowledge graphs and hybrid search. 17+ AI tools via MCP. Free.5% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.231 GitHub stars.
AI agent governance MCP servers
A server here enforces policy on, approves, audits or restricts what AI agents are allowed to do.
Not here: servers that are merely used by agents.
733 remote1,124 localTagged by Odel's AI, which can make mistakes
Most popular in AI agent governance
By GitHub stars, at most two per publisher- Number 1: Claude Flow@ruvnet/claude-flowAI orchestration with hive-mind swarms, neural networks, and 87 MCP tools for enterprise dev.9% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.73,814 GitHub stars.
- Number 2: Ouroboros@q00/ouroborosPins an acceptance spec; the verify command and expected output never enter the success contract.10% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from PyPI.6,192 GitHub stars.
- Number 3: Tradememory Protocol@mnemox-ai/tradememory-protocolTamper-evident decision audit trail and outcome-weighted memory for AI trading agents.6% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from PyPI.1,427 GitHub stars.
