LMCP Tool Builder

A Python library for discovering, building and loading tools from LMCP (Language Model Context Protocol) servers.

Features

• Tool Discovery: Automatically discover tools from LMCP servers
• Local Caching: Cache tools locally for offline use
• Auto Update: Configurable auto-update behavior
• Tool Testing: Built-in testing functionality for discovered tools
• Easy Integration: Simple API for integrating with your applications

Installation

pip install lmcp-tool-builder

Quick Start

from lmcp_tool_builder import LMCPToolBuilder
import os

# Create a tool builder instance
builder = LMCPToolBuilder(
    server_url="http://your-lmcp-server:8000",
    api_key=os.getenv("LMCP_API_KEY"),
    local_tools_file="bot_tools.py",
    debug=True,
    auto_update=True
)

# Build and load tools
tools = builder.build_and_load_tools()

# Use the tools
for tool in tools:
    print(f"Tool: {tool.__name__}")

Configuration

Environment Variables

Create a .env file in your project:

LMCP_API_KEY=your_api_key_here
LMCP_LOCAL_TOOLS_FILE=bot_tools.py

LMCPToolBuilder Parameters

• server_url: URL of the LMCP server (default: "http://localhost:8000")
• api_key: API key for authentication (default: "")
• local_tools_file: Path to local tools cache file (default: "bot_tools.py")
• debug: Enable debug output (default: False)
• auto_update: Auto-update tools from server (default: False)

API Reference

LMCPToolBuilder Class

Methods

• discover_tools(): Discover tools from LMCP server or local cache
• build_tools_module(server_tools): Build a Python module from server tools
• save_tools_module(module_code): Save module to local file
• load_tools_from_module(): Load tools from local module
• build_and_load_tools(): Main method to build and load tools
• test_tools(tools): Test loaded tools with sample inputs

Example Usage

from langchain_openai import ChatOpenAI
from langchain.agents import create_agent
from lmcp_tool_builder import LMCPToolBuilder # lmcp_tool_builder
import asyncio

import os
from dotenv import load_dotenv

load_dotenv()

# 到lmcp平台注册创建应用、绑定工具、获得api_key（略）
print(f'[LMCP] 开始更新工具...')

# 创建工具构建器
builder = LMCPToolBuilder(
    server_url="http://aicity.wang:8000",
    api_key=os.getenv("aca8942ed1b43ab16f796eb223db10dc12edaa0c3fad7d41b85215bcb6aceefb"),
    local_tools_file=os.getenv("LMCP_LOCAL_TOOLS_FILE", "bot_tools.py"),
    debug=True,
    auto_update=False
)

# 构建并加载工具
tools = builder.build_and_load_tools()

# 构建agent
agent = create_agent(
    model=ChatOpenAI(
        model="deepseek-chat",
        api_key="your api key",
        base_url="https://api.deepseek.com/v1"
    ),
    tools=tools
)


# 异步流式输出
async def async_stream_agent():
    async for message, metadata in agent.astream(
        {"messages": [("user", "你必须使用工具，异步计算2+8=？告诉我你使用哪工具")]}, 
        stream_mode="messages"
    ):
        # 实时打印 Token
        if message.content:
            print(message.content, end="", flush=True)

# 同步流式输出
def stream_agent():
    for token, metadata in agent.stream(
        {"messages": [{"role": "user", "content": "1+2=?"}]},
        stream_mode="messages" #update
    ):
        if token.content:
            print(token.content, end="", flush=True)


if __name__ == "__main__":
    stream_agent()
    asyncio.run(async_stream_agent())

Development

Setup Development Environment

# Clone the repository
git clone https://github.com/example/lmcp-tool-builder.git
cd lmcp-tool-builder

# Install in development mode
pip install -e .

Running Tests

python -m pytest tests/

License

MIT License - see LICENSE file for details.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.