CRM Manager MCP Server

A Model Context Protocol (MCP) server that provides an interface to manage contacts, leads, and deal stages in a CRM environment. Built with the FastMCP framework, this server allows AI assistants (like Claude) to search your database, create leads, and track revenue stats in real-time.

🚀 Features

Tools

Tools allow the AI to perform actions or fetch specific data based on user intent.

• Contacts: Search the database by name or email.
• create_lead: Programmatically add new entries to the contact list.
• update_deal_stage: Modify the status of active deals (e.g., moving from "Negotiation" to "Closed Won").

Resources

Resources provide the AI with read-only context or "knowledge" documents.

• crm://contacts/all: A full dump of the contact database.
• crm://stats: A high-level summary of lead counts and total revenue.

🛠 Installation

1. Clone the repository:

git clone <your-repo-url>
cd crm-server

1. Install dependencies: Ensure you have Python 3.10+ installed.

pip install mcp

💻 Usage

Running Locally

To start the server manually for testing:

python crm_server.py

Claude Desktop Integration

To use this with Claude Desktop, add the following to your claude_desktop_config.json:

{
  "mcpServers": {
    "crm-manager": {
      "command": "python",
      "args": ["/absolute/path/to/your/crm_server.py"]
    }
  }
}

📊 Data Schema

The server currently operates on a mock in-memory database with the following structure:

🧪 Example Prompts

Once connected, you can ask your AI assistant:

• "Search for Alice in the CRM."
• "Add a new lead: John Doe at john@doe.com."
• "What is our total revenue right now?" (This triggers the crm://stats resource).
• "Mark deal #1 as Closed Won."

⚖️ License

MIT