mcpField
M
MCP Bone

MCP Bone

Local
github.com/tangshuangAI & Agents0Updated 79w ago
Repository

About MCP Bone

A Node.js MCP server that connects to MCP Bone, an online service for registering MCP servers and obtaining tool definitions. It also provides parsers that convert completion text into tool-call JSON, including for LLMs without function calling support.

Key features

  1. 01Connects to MCP Bone
  2. 02Retrieves tool JSON
  3. 03Parses text into tool-call JSON
  4. 04Supports JSON and XML-style tool formats

Use cases

  1. 01Fetching registered tool definitions from MCP Bone
  2. 02Parsing LLM completion text without function calling
  3. 03Converting XML-style tool text into JSON

MCP support

The MCP server connects to MCP Bone and retrieves tool JSON. MCP Bone lets users register other MCP servers and obtain their tool definitions.

#mcp#mcp-server#mcp-bone

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

mcp.json · Universal config
{
  "mcpServers": {
    "bone": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-bone"
      ],
      "env": {
        "MCP_BONE_ACCESS_TOKEN": "<MCP_BONE_ACCESS_TOKEN>"
      }
    }
  }
}
Run commandnpx -y mcp-bone
Environment variables to set
MCP_BONE_ACCESS_TOKENSecret

Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.

Packages
npmmcp-bone

Tools

No tool list yet

Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.

Details

From the project README

MCP Server of MCP Bone

What's MCP Bone?

MCP Bone is a online service that provides a set of tools for MCP. Home Page On MCP Bone, you can register other MCP Servers, get function calling tools JSON or XML-style prompt text and later call tools easily. Read Usage

This package is a Node.js module that provides a MCP Server connect to MCP Bone. You can use it to connect to MCP Bone and get the tools JSON.

Also, you can use this package as a parser SDK to parse completion text to tool_calls.

Setup

Access Token

Get an Access Token:

  1. go to https://store.tangshuang.net?scope=mcp-bone
  2. register an account and log in
  3. go to https://store.tangshuang.net/product/58/entry?scope=mcp-bone to create a new App
  4. go to the App's Certificate menu page to create a new certificate
  5. copy the Access Token
NPX
{
    "mcpServers": {
        "mcp-bone": {
            "command": "npx",
            "args": [
                "-y",
                "mcp-bone"
            ],
            "env": {
                "MCP_BONE_ACCESS_TOKEN": ""
            }
        }
    }
}
Parser
import {
    extractAndParseToolsToJson,
    extractAndParseToolsXmlToJson,
    extractAndParseToolsTextToJson,
} from 'mcp-bone';

const tools = extractAndParseToolsTextToJson(text);
const tools = extractAndParseToolsXmlToJson(text);
const tools = extractAndParseToolsToJson(text);

These tools are used after LLM completions whic does not support function calling to parse the text into tools JSON. If the result is an empty array, it means that the LLM did not generate any tools.

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.