Enterprise Model Context Protocol (MCP) Server & Client

Name: Enterprise Model Context Protocol (MCP) Server & Client
Author: sanjay-si

@sanjay-si

Repository

About Enterprise Model Context Protocol (MCP) Server & Client

MCP Client & Server

#enterprise-mcp #model-context-protocol #llm-integration

Basic information

Config

Add this server to your MCP-compatible client using the configuration below.

{
  "mcpServers": {
    "enterprise_mcp": {
      "command": "python",
      "args": [
        "-m",
        "src.mcp_server.main"
      ]
    }
  }
}

Tools

No tools detected

We auto-extract tools from the README. The maintainer can list them under a ## Tools heading to populate this section.

Overview

What is Enterprise Model Context Protocol (MCP) Server & Client?

A comprehensive enterprise-grade implementation of the Model Context Protocol (MCP) for connecting LLMs with enterprise tools and data sources. It is cost-optimized for production, allows zero-cost testing without any LLM API keys, and supports either OpenAI GPT or Anthropic Claude with a single API key.

How to use Enterprise Model Context Protocol (MCP) Server & Client?

Clone the repository, then start all services with Docker Compose or manually install server and client components. Set environment variables (choose one LLM API key or none for zero-cost testing) and run the server (e.g., python -m src.mcp_server.main) and client (e.g., python examples/demo.py or python examples/demo_no_llm.py). Connect the client to the server via WebSocket and use the provided Python examples to invoke enterprise tools.

Key features of Enterprise Model Context Protocol (MCP) Server & Client

Full JSON‑RPC 2.0 MCP specification compliance
Enterprise tools: database queries, file operations, API integrations
JWT authentication and role‑based access control
Prometheus metrics and structured logging
Multi‑LLM support (OpenAI GPT, Anthropic Claude)
Async/await architecture and robust error handling
Zero‑cost testing mode with no API keys required
Optimized models (gpt‑4o‑mini, claude‑3.5‑haiku) for 97% cost savings

Use cases of Enterprise Model Context Protocol (MCP) Server & Client

Query an enterprise database and generate a report file via one LLM request
Perform sandboxed file system operations (read, write, create) through MCP tools
Make authenticated HTTP requests to external enterprise APIs
Enable secure, monitored access to enterprise data for AI assistants

FAQ from Enterprise Model Context Protocol (MCP) Server & Client

What runtime dependencies are required?

Python 3.11+ is required. PostgreSQL and Redis are optional – only needed for enterprise database features and caching/session management, respectively.

Can I test the system without any LLM API keys?

Yes. The implementation supports zero‑cost testing that works without any LLM API keys. You can run python examples/demo_no_llm.py to try enterprise tools without connecting to an LLM.

How do I configure authentication and security?

Set JWT_SECRET_KEY, JWT_ALGORITHM, and JWT_EXPIRATION_HOURS in your environment. The server enforces JWT authentication and role‑based access control, with SQL injection protection and sandboxed file access.

What transport does the MCP server use?

The server communicates over WebSocket. The client connects to ws://localhost:8000/mcp by default. HTTPS/WSS is supported for encrypted communications.

What monitoring is available?

Prometheus metrics are exposed at GET /metrics, including request counts, request duration, active connections, and tool execution statistics. A basic health check endpoint is also provided at GET /health.