MCP Enterprise Server
by vivekjais16
README.md
# MCP Enterprise Server (Model Context Protocol Gateway)
[](https://github.com/vivekjais16/mcp-enterprise-server/actions)
[](https://www.python.org/downloads/)
[](https://fastapi.tiangolo.com)
[](https://modelcontextprotocol.io/)
**Author:** Vivek Jaiswal (<vivekjais16@gmail.com>)
**Role:** Senior Software Engineer — Python | Django | FastAPI | Generative AI & Agentic AI
---
## 🏛️ Architectural Overview
**MCP Enterprise Server** is a high-performance, security-hardened implementation of Anthropic's **Model Context Protocol (MCP)** specification (JSON-RPC 2.0). It provides stateful AI agents, LLM tool-calling engines, and client interfaces (Claude Desktop, LangGraph supervisors, Cursor) with real-time access to production tools, enterprise databases, knowledge bases, and system observability metrics.
```
+-------------------------------------------------------------+
| LLM Client / Autonomous Agent |
| (Claude Desktop / LangGraph Router) |
+------------------------------+------------------------------+
|
JSON-RPC 2.0 | SSE Stream / Stdio
v
+-------------------------------------------------------------+
| MCP Enterprise Server (FastAPI) |
| +---------------------+ +-------------------------------+ |
| | Bearer Auth & RBAC | | Rate Limiter (120 req/min) | |
| +---------------------+ +-------------------------------+ |
| +--------------------------------------------------------+ |
| | JSON-RPC 2.0 Engine | |
| | (initialize, tools/*, resources/*, prompts/*) | |
| +--------------------------------------------------------+ |
| | | | |
| v v v |
| +---------------+ +------------+ +-------------------+ |
| | Database SQL | | Knowledge | | Sandboxed Python | |
| | Safety Guard | | Search RRF | | AST Inspector | |
| +---------------+ +------------+ +-------------------+ |
+-------------------------------------------------------------+
```
---
## ⚡ Core Capabilities
### 1. Enterprise Tools
- **`query_database`**: Parameterized SQL query execution with AST validation that blocks mutations (`DROP`, `DELETE`, `UPDATE`, `ALTER`) in read-only agent contexts.
- **`search_knowledge_base`**: Hybrid semantic retrieval across indexed technical documentation with vector scoring.
- **`system_metrics`**: Live host CPU, RAM, disk, and process telemetry inspection for automated SRE agents.
- **`execute_code_sandbox`**: Isolated Python execution with AST import inspection, blocking unsafe primitives (`os`, `sys`, `eval`, `__import__`).
### 2. Contextual Resources
- `system://status` — Live operational health and worker telemetry.
- `database://schema` — Relational schema reflection and column definitions.
- `architecture://overview` — Enterprise microservices topology.
### 3. Prompt Templates
- `audit_sql_query` — Pre-built prompt structure for SQL index optimization and injection audits.
- `incident_investigation` — Automated incident response runbook generation.
### 4. Transports
- **Server-Sent Events (SSE)** (`/sse`): High-throughput streaming transport for distributed microservices and web agents.
- **Stdio Transport** (`src/transports/stdio.py`): Standard I/O interface compatible with Claude Desktop and command-line LLM runners.
---
## 🚀 Quick Start
### 1. Clone & Setup
```bash
git clone https://github.com/vivekjais16/mcp-enterprise-server.git
cd mcp-enterprise-server
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt
```
### 2. Run Test Suite
```bash
pytest tests/ -v
```
### 3. Launch Server
```bash
uvicorn src.main:app --host 0.0.0.0 --port 8080 --reload
```
---
## 🔒 Security & Authorization
All requests to direct JSON-RPC (`/mcp/rpc`) and SSE (`/sse`) require Bearer token or API key authentication:
```bash
curl -X POST http://localhost:8080/mcp/rpc \
-H "Authorization: Bearer mcp_live_secret_key_prod_8920171244" \
-H "Content-Type: application/json" \
-d '{
"jsonrpc": "2.0",
"id": 1,
"method": "tools/list"
}'
```
---
## 🐳 Docker Deployment
```bash
docker-compose up -d --build
```
This server cannot be deployed
Maintenance
ActivityMaintained
ResponsivenessNo issues