invoice-mcp-system
Multi-component AI-powered system for invoice processing, document analysis, and business automation with integrated MCP servers for Claude Code
GitHubスター
0
ユーザー評価
未評価
フォーク
0
イシュー
0
閲覧数
1
お気に入り
0
Claude MCP Servers
This project demonstrates the integration of Model Context Protocol (MCP) servers with Claude Desktop.
Available MCP Servers
1. Fetch MCP Server
The MCP Fetch server provides web content fetching capabilities, enabling Claude to retrieve and process content from web pages.
"fetch": {
"command": "node",
"args": ["/Users/andrew/Projects/claudecode1/src/fetch/server.js"]
}
2. Playwright MCP Server
The Playwright MCP server enables browser automation capabilities, allowing Claude to interact with web pages.
"playwright": {
"command": "npx",
"args": ["@playwright/mcp@latest", "--config", "/Users/andrew/Projects/claudecode1/src/playwright/config/playwright-mcp-config.json"]
}
3. Vision OpenAI MCP
A document processing service that uses OpenAI's Vision models to analyze images and PDFs, with special handling for invoices.
"vision": {
"type": "stdio",
"command": "python",
"args": ["/Users/andrew/Projects/claudecode1/src/vision-openai-mcp/server.py"]
}
4. CRM Database MCP Server
A comprehensive CRM system with contact management, communication processing (email, iMessage, WhatsApp, etc.), and financial tracking. Designed specifically for Claude Code with LLM-optimized tools and resources.
Features:
- Universal communication tracking (email, SMS, WhatsApp, iMessage, Instagram DMs, physical mail)
- Contact and company relationship management
- Financial tracking (accounts, transactions, subscriptions)
- Task management and follow-up automation
- Cross-platform identity resolution
- AI-ready data structure with built-in processing fields
"crm-db": {
"command": "python3",
"args": ["/Users/andrew/Projects/claudecode1/src/crm-db/server.py"],
"env": {
"CRM_DB_PATH": "/Users/andrew/Projects/claudecode1/src/crm-db/crm.db"
}
}
Setup Vision OpenAI MCP
Install dependencies:
pip install "mcp[cli]" httpx python-dotenv pdf2image Pillow
Set your OpenAI API Key:
cd src/vision-openai-mcp cp .env.example .env # Edit .env to add your OpenAI API key
Use the server in Claude Desktop by adding it to your configuration.
Setup CRM Database MCP
Install dependencies:
cd src/crm-db pip install -r requirements.txt
Initialize database:
python database.py --init --sample-data
Test the server:
python test_server.py
The server is automatically available in Claude Code via the
mcp.json
configuration.
Running the Servers
You can run the servers directly using the scripts in package.json:
# Run Fetch MCP
npm run mcp-fetch
# Run Playwright MCP
npm run mcp-playwright
For the Vision MCP, you can run it with:
python src/vision-openai-mcp/server.py
For the CRM Database MCP:
python3 src/crm-db/server.py
Note: The CRM server is designed for use with Claude Code and will be automatically started when Claude Code initializes the MCP servers.
Requirements
- Claude Code (for CRM MCP Server)
- Claude Desktop (for other servers)
- Python 3.8+ with dependencies
- Node.js and npm for Playwright MCP
- OpenAI API key for Vision MCP
Documentation
For more detailed information about each MCP server, see the individual READMEs in their respective directories:
CRM Database MCP Additional Guides
- Testing Guide - Comprehensive testing instructions
- Claude Code Usage Guide - How to use with Claude Code