3GPP

You run an MCP server that lets AI assistants search and retrieve information from 3GPP specification documents. It bundles pre-processed data, supports full-text search, and exposes tools to query documents, look up NAS/GMM/IMSCause values, and list available specifications. This guide shows you how to use the server, install it locally, and configure it with MCP clients.

How to use

You interact with the MCP server through a client that issues structured requests. Use the search tool to find specifications by keywords, ask for EMM/5GMM causes, or request lists of available specifications. The server returns structured results you can render in your UI and pass to your AI agent for user-facing responses.

Typical usage patterns include searching for a topic like "Search 3GPP documents for attach procedure" or retrieving a specific NAS/GMM/IMSCause value such as "EMM cause #3" or "5GMM cause #7". You can also request a list of all specifications provided by the server to understand what is available.

How to install

Prerequisites: ensure you have Node.js installed (version 18.0.0 or newer). You may also need Git LFS if you plan to pull pre-built data stored with large files.

Step 1: Install Git LFS if you plan to use pre-built data Run the following command to enable Git LFS on your system:

git lfs install

Step 2: Install the MCP server locally by cloning the project and installing dependencies.

git clone https://github.com/Lee-SiHyeon/mcp-server-3gpp.git
cd mcp-server-3gpp
npm install

Step 3: Pull pre-built data if you are using Git LFS data files.

git lfs pull

Step 4: Start the MCP server after dependencies are installed.

npm start

Configuration and MCP client setup

Configure your MCP client to point at the local stdio server or to a remote MCP service if you have one. The examples below show how to wire a local stdio server using Node.js and the provided entry point.

{
  "servers": {
    "3gpp_docs": {
      "type": "stdio",
      "command": "node",
      "args": ["src/index.js"]
    }
  }
}

If you prefer to configure the server for a client named 3gpp_docs in Claude Desktop, you can use the following structure which is functionally equivalent for local usage.

{
  "mcpServers": {
    "3gpp_docs": {
      "command": "node",
      "args": ["src/index.js"]
    }
  }
}

You can also pass a custom environment variable to specify the chunks data path if you keep the data file in a non-default location.

{
  "servers": {
    "3gpp_docs": {
      "type": "stdio",
      "command": "node",
      "args": ["src/index.js"],
      "env": {
        "CHUNKS_FILE_PATH": "/path/to/your/chunks.json"
      }
    }
  }
}

Notes and troubleshooting

If you are not using Git LFS for large data files, the server will still function for EMM/5GMM cause lookups but full document search relies on the pre-built data file. Ensure the chunks.json data file is available at the path specified in env variables if you customize the data location.

If you encounter issues starting the server, verify Node.js is installed and that npm install completed successfully. Check that the src/index.js path exists in your project and that any environment variables referenced in the config are set correctly.

Usage and tooling overview

The MCP server provides three core tools you can use from your AI client: search_3gpp_docs to search 3GPP documents by keywords, get_emm_cause to retrieve EMM/5GMM cause details, and list_specs to enumerate available specifications.

Available tools

search_3gpp_docs

Search 3GPP documents by keywords and retrieve relevant results in a structured response.

get_emm_cause

Get EMM cause details for LTE or 5GMM causes for 5G.

list_specs

List all available specifications included in the MCP server data set.