Salesforce数据自然语言查询桥接器

# 1. Clone the repository
git clone https://github.com/yourusername/mcp-openai-bridge.git
cd mcp-openai-bridge

# 2. Install dependencies
pip install -r requirements.txt

# 3. Set environment variables
export OPENAI_API_KEY="sk-your-openai-key"
export SF_CLIENT_ID="your-salesforce-client-id"
export SF_CLIENT_SECRET="your-salesforce-client-secret"
export SERVER_PATH="/path/to/your/mcp-server/server.py"
export PYTHON_PATH="/usr/local/bin/python3"  # Optional, adjust to your Python path

python interactive_chat.py

🚀 MCP-OpenAI Interactive Chat
============================================================
🔌 Connecting to MCP server...
✅ Connected! Found 3 tools

💬 You: What tables are available in my Data 360?

🤖 Assistant: I found the following tables:
- Customer__dlm
- Orders__dlm  
- Products__dlm

💬 You: Show me the top 5 customers by revenue

🤖 Assistant: Here are your top 5 customers by revenue:
1. Acme Corp - $1.2M
2. TechStart Inc - $980K
...

streamlit run streamlit_app.py

User Input → OpenAI GPT-4 → MCP Bridge → MCP Server → Salesforce Data 360

# Required
OPENAI_API_KEY=sk-your-openai-api-key
SF_CLIENT_ID=your-salesforce-client-id
SF_CLIENT_SECRET=your-salesforce-client-secret
SERVER_PATH=/path/to/mcp-server/server.py

# Optional
PYTHON_PATH=/usr/local/bin/python3
OPENAI_MODEL=gpt-4o

import asyncio
from mcp_openai_bridge import MCPOpenAIBridge

async def main():
    bridge = MCPOpenAIBridge(
        openai_api_key="sk-your-key",
        mcp_command="python3",
        mcp_args=["/path/to/server.py"],
        mcp_env={
            "SF_CLIENT_ID": "your-id",
            "SF_CLIENT_SECRET": "your-secret"
        }
    )
    
    await bridge.connect()
    response = await bridge.chat("What tables do I have?")
    print(response)
    await bridge.close()

asyncio.run(main())

pip install -r requirements.txt