Gatekeeper MCP Server

Run MCP tool calls through Runestone Gatekeeper — policy enforcement, human approval, audit.

Local serverstdioPython

What is the Gatekeeper MCP server?

Gatekeeper MCP server exists for a simple reason — assistants are far more useful when they can act on Gatekeeper directly instead of describing what you should do. Run MCP tool calls through Runestone Gatekeeper — policy enforcement, human approval, audit.

What you get

A policy-based gatekeeper service that sits between AI agents and real-world tools (shell, HTTP, filesystem), enforcing approvals, denials, USD budgets, and audit logging.

Setting it up

Installation goes through your MCP client rather than a global install: point it at @runestone-labs/gatekeeper-claude-code on npm and it is fetched when the client starts. The copy-paste blocks for Claude Desktop, Claude Code and Cursor are further down this page.

What the assistant can call

Once Gatekeeper is connected, these are the calls the assistant has available:

  • memory.upsert — Create/update entities (people, projects, concepts)
  • memory.link — Create relationships between entities
  • memory.unlink — Remove relationships between entities
  • memory.query — Query entities (with full-text search) and traverse relationships
  • memory.episode — Log decisions, events, and observations
  • memory.evidence — Attach evidence/provenance to entities or episodes
  • Recording — The Recording tool exposed by this server
  • Outputs — The Outputs tool exposed by this server
  • Configuration — The Configuration tool exposed by this server
  • Guides — The Guides tool exposed by this server
  • Reference — The Reference tool exposed by this server
  • Contributing — The Contributing tool exposed by this server

Configuration and credentials

You will need 8 environment variables: GATEKEEPER_URL, GATEKEEPER_ROLE, GATEKEEPER_SECRET, APPROVAL_PROVIDER, AUDIT_SINK, POLICY_SOURCE, SLACK_WEBHOOK_URL, GATEKEEPER_PORT. The server will not start without them, which is usually why the tools fail to appear on a first run. Keep credentials in your client's env block or a secrets manager rather than in a file you might commit.

Before you rely on it

  • It runs with your machine's permissions. That is convenient and also the reason to think about what you point it at before you approve a tool call.
  • With 12 tools registered it takes up a noticeable share of the context window; turn it off in projects that never touch Gatekeeper.
  • Missing credentials fail quietly in some clients — if no tools show up, check the environment block first.
  • MCP clients confirm each tool call by default. Leave that on until you have watched what the gatekeeper mcp server does with a few real requests.

Choosing this one

Among the knowledge and memory options, the useful question is rarely "what can it do" but "what does it cost you to run" — permissions, credentials, and how much of your context its toolset consumes. Gatekeeper's toolset — memory.upsert, memory.link, memory.unlink and 9 more — is a fair guide to whether it matches your workflow. It is maintained by Runestone-Labs; worth a glance at recent repository activity before you build anything load-bearing on it.

SyncDev reviews every entry in this directory against the project's own documentation before publishing, and revisits them as servers change.

Available tools

ToolWhat it does
memory.upsertCreate/update entities (people, projects, concepts)
memory.linkCreate relationships between entities
memory.unlinkRemove relationships between entities
memory.queryQuery entities (with full-text search) and traverse relationships
memory.episodeLog decisions, events, and observations
memory.evidenceAttach evidence/provenance to entities or episodes
RecordingThe Recording tool exposed by this server.
OutputsThe Outputs tool exposed by this server.
ConfigurationThe Configuration tool exposed by this server.
GuidesThe Guides tool exposed by this server.
ReferenceThe Reference tool exposed by this server.
ContributingThe Contributing tool exposed by this server.

How to install the Gatekeeper MCP server

{
  "mcpServers": {
    "gatekeeper": {
      "command": "npx",
      "args": ["-y", "@runestone-labs/gatekeeper-mcp"],
      "env": {
        "GATEKEEPER_URL": "http://127.0.0.1:3847",
        "GATEKEEPER_ROLE": "agent"
      }
    }
  }
}

Configuration as documented by the project. Restart the client after saving.

Configuration

VariableDescriptionRequired
GATEKEEPER_URLEndpoint or connection string the server talks to.Yes
GATEKEEPER_ROLEConfiguration value read at startup.Optional
GATEKEEPER_SECRETCredential the server authenticates with.Yes
APPROVAL_PROVIDERConfiguration value read at startup.Optional
AUDIT_SINKConfiguration value read at startup.Optional
POLICY_SOURCEConfiguration value read at startup.Optional
SLACK_WEBHOOK_URLEndpoint or connection string the server talks to.Yes
GATEKEEPER_PORTConfiguration value read at startup.Optional

Example prompts to try

  • Use Gatekeeper to memory.upsert.
  • Use Gatekeeper to memory.link.
  • Use Gatekeeper to memory.unlink.

Frequently asked questions

It connects Gatekeeper to MCP-compatible AI assistants such as Claude and Cursor, exposing 12 tools (memory.upsert, memory.link, memory.unlink, and more) that the assistant can call on your behalf. Instead of copying data back and forth by hand, the assistant works with Gatekeeper directly.