Skip to content
MCP server

HOL Guard

Review and block risky AI agent actions, installs and MCP calls on your machine

Runs locallyStatus and history
Install
Pricing
Freemium
Platforms
CLI, self-hosted
Runs
Locally
Version
3.59.0
Listing updated
3 days ago

About HOL Guard

HOL Guard is a freemium MCP server for security. It runs locally on your machine and is started by your AI client.

Guard sits between an agent and your computer and applies a policy you choose before a covered action runs. It looks at shell commands, file access, package installs, MCP tool calls and prompts, and flags secret exposure, destructive operations, prompt injection and supply-chain risks, then allows, asks or blocks. Coverage depends on the integration, and it records receipts for later review.

The MCP server is a stdio part of the hol-guard Python package, which also provides a CLI and a local dashboard. Setup is pipx install hol-guard and a first-run wizard that discovers supported agents and asks before each change. It runs without an account, and the site says files are not uploaded.

Guard Cloud is separate and optional: Solo keeps decision history across devices and Pro and Team add shared policy. A related plugin scanner can check extensions in maintainer CI. The project is developed by HOL, which describes itself as a consortium building agentic AI standards.

Install

Run in a terminal

claude mcp add hol-guard -- uvx hol-guard

Add to claude_desktop_config.json

{
  "mcpServers": {
    "hol-guard": {
      "command": "uvx",
      "args": [
        "hol-guard"
      ]
    }
  }
}
Install in CursorOpens Cursor, which asks you to confirm. Or add it by hand:

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "hol-guard": {
      "command": "uvx",
      "args": [
        "hol-guard"
      ]
    }
  }
}
Install in VS CodeOpens VS Code, which asks you to confirm. Or add it by hand:

Add to .vscode/mcp.json

{
  "servers": {
    "hol-guard": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "hol-guard"
      ]
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "hol-guard": {
      "command": "uvx",
      "args": [
        "hol-guard"
      ]
    }
  }
}

Add to your client's MCP settings (most use this shape)

{
  "mcpServers": {
    "hol-guard": {
      "command": "uvx",
      "args": [
        "hol-guard"
      ]
    }
  }
}

What it does

Allow, ask or block
Decisions are made before a covered action runs.
Broad coverage
Shell, files, installs, MCP tool calls and prompts.
Local dashboard
Approve actions and review decision history.
Package screening
Evaluates supported package installs against supply-chain data.
Plugin scanner
Validates extensions in CI before release.
Optional cloud
Shared history and team policy as a separate plan.

Good for

  • Stop an agent from reading cloud credentials
  • Require approval before destructive commands
  • Review a new MCP server before it launches
  • Keep an audit trail of agent decisions
  • Gate plugin releases with a scanner

Pricing

Freemium

Free local use; optional Guard Cloud (Solo, Pro, Team) prices not published

Prices change. Check hol.org before you buy.

HOL Guard FAQ

Do I need an account?

No, it runs locally without one.

How do I install it?

pipx install hol-guard, then hol-guard init.

Does it upload my files?

The site says files never leave your machine.

Is there a paid tier?

Guard Cloud is optional; prices are not shown.

Is HOL Guard free?

HOL Guard has a free plan, with paid plans for more usage or features.

How do I add HOL Guard to Claude Code?

Run this command in your terminal: claude mcp add hol-guard -- uvx hol-guard

What are the best alternatives to HOL Guard?

The closest alternatives to HOL Guard are REA, Agent Roast, Midplane, Leash and Tinplate Edge AI Safety, Security & EU Compliance.

Alternatives to HOL Guard

  1. REA

    MCP server

    CLI and MCP server that lets coding agents reverse engineer apps and binaries

    Runs locally
    • Open source
    • 464K GitHub stars
    • Developer tools
  2. Agent Roast

    MCP server

    Offline review that flags agent changes which weaken tests, assertions or CI

    Runs locally
    • Open source
    • 4 Hacker News points
    • Developer tools
  3. Midplane

    MCP server

    Postgres gateway for AI agents with table grants, column masking and approval for writes

    Runs locally
    • Open source
    • 5 GitHub stars
    • Databases
  4. Leash

    MCP server

    Review what your AI touched, approve risky actions and undo supported changes

    • 1 tool
    • Remote
    • Free
    • 1 GitHub star
    • Security
  5. SecureCloud

    MCP server

    SentinelOne and IT security catalog, prices, quotes and checkout links for Israel

    • 7 tools
    • Remote
    • Free
    • Security