MCP Server product markMODEL CONTEXT PROTOCOL

The Official LabelGrid MCP Server

Connect Claude, Cursor, or any AI assistant to your catalog and run your music distribution in plain language.

The LabelGrid MCP server is our first-party, open-source connector for the Model Context Protocol. Point Claude Desktop, Claude Code, Cursor, or any MCP-compatible client at your LabelGrid account, then create releases, check review status, and pull analytics by asking for what you want. It ships 80+ tools across ten toolsets and runs as a thin, typed wrapper over the LabelGrid public API.

Get It on npm

View the source on GitHub →

Chat bubble streaming a ribbon into a waveform core — natural language driving music distribution

SIXTY-SECOND SETUP

Add one line. Your AI walks you through the rest.

The server runs through npx, so there is nothing to download and nothing to keep updated. Add one entry to your MCP client’s config and you are connected:

{
  "mcpServers": {
    "labelgrid": {
      "command": "npx",
      "args": ["-y", "@labelgrid/mcp"],
      "env": {
        "LABELGRID_API_TOKEN": "YOUR_API_TOKEN"
      }
    }
  }
}

Start the server without a token and it drops into a guided setup mode: your AI assistant walks you through connecting your account, step by step. Generate a token under Profile → API Tokens in your dashboard, drop it into the config, and you are working your catalog in natural language. Node.js 20 or newer is the only requirement.

IN YOUR OWN WORDS

Ask for it, and it happens

These are the kinds of requests the server handles. Each one maps to a real toolset, so your assistant isn’t guessing — it’s calling the same API the dashboard uses.

Releases & tracks

“Draft a twelve-track album under my main label and set the release date to next Friday.”

The AI creates the draft release and its tracks, ready for you to review before anything goes out.

Review & quality

“Why is my Midnight Sessions release stuck in review?”

It reads the open review issues and tells you exactly what needs fixing — and can check the optional Stream Radar early-warning flags while it’s at it.

Accounting & royalties

“Pull last quarter’s royalty statement and break it down by label.”

It fetches your statements and splits the numbers the way you asked, no spreadsheet export required.

Analytics

“Show streams for my latest single by store this week.”

It pulls your streaming analytics and breaks the numbers down the way you asked, per store and per period.

Webhooks

“Add a webhook that notifies my server whenever a delivery completes.”

It configures the webhook against your account so your own systems stay in sync automatically.

Delivery status

“Which stores has the single gone live on, and is anything still pending?”

It checks delivery status across your destinations and reports what is live and what is still in flight.

WORKS WITH

Any MCP client, open source end to end

Claude Desktop, Claude Code, Cursor, and any other client that speaks the Model Context Protocol connect the same way. The server is first-party and open source under the MIT license, so you can read every line, fork it, or open an issue. It is built on the open Model Context Protocol standard, which any compatible assistant can talk to.

Star it on GitHub →

BUILT TO BE TRUSTED

Three permission tiers, fail-closed

Reads — always on

Look up releases, review issues, analytics, royalty statements, and delivery status. Read access is available on every connection, so your assistant can answer questions from day one.

Safe writes — on by default

Draft releases and tracks, edit metadata, manage labels, artists, writers and publishers, set up webhooks, and build smart-link landing pages. You can switch these off or force the whole connection read-only whenever you want.

Consequential actions — off by default

Distribution submission, takedowns, and immutable audio and artwork uploads stay disabled until you explicitly turn them on. Turning them on takes an opt-in plus a typed acknowledgment before anything can run.

Every rule and validation runs on LabelGrid’s servers, so the AI can never bypass your account’s protections — the same checks that apply in the dashboard apply here. To enable consequential actions you opt in and then type the exact acknowledgment “I accept responsibility for AI-driven distribution actions” before any AI-driven submission or takedown will run. Distribution submissions still count against your weekly release limits.

GETTING ACCESS

Included with LabelGrid API access

The MCP server talks to the LabelGrid public API, so it is available on LabelGrid’s API plans. Already have API access? Generate a token and connect today. New to the API? Start with the API overview, then choose a plan that fits.

Read the API overview →  ·  For developers →

QUESTIONS

MCP server FAQ

The LabelGrid MCP server is our official, open-source connector that links an AI assistant to your LabelGrid account through the Model Context Protocol. It exposes 80+ tools across ten toolsets, so you can create releases, check review status, pull analytics and royalties, and manage delivery by asking in plain language. It is a thin, typed wrapper over the LabelGrid public API.

The Model Context Protocol is an open standard for connecting AI assistants to external tools and data. An MCP server exposes a set of actions an AI client can call, so an assistant like Claude or Cursor can work with a service directly. LabelGrid publishes an MCP server so your assistant can operate your distribution account, and you can read the specification at modelcontextprotocol.io.

Yes. The MCP server connects through the LabelGrid public API, which is part of LabelGrid’s API plans. If you already have API access you can generate a token and connect right away, and if you don’t, the API overview and the pricing page will get you started.

No. Distribution submissions, takedowns, and immutable uploads are off by default and stay disabled until you explicitly opt in and type an acknowledgment confirming you accept responsibility for AI-driven distribution actions. Reads and safe draft edits are available without that step, and every action is still validated on LabelGrid’s servers before it takes effect.

Any client that supports the Model Context Protocol works, including Claude Desktop, Claude Code, and Cursor. You add one entry to the client’s configuration and the server handles the rest, walking you through setup if you start it without a token.

Yes. The LabelGrid MCP server is first-party and open source under the MIT license. The code is on GitHub and the package is published on npm as @labelgrid/mcp, so you can inspect it, contribute, or report an issue.

START NOW

Put your catalog one message away

Connect your AI assistant to LabelGrid and run releases, analytics, and royalties in plain language.

Read the API overview →