home / mcp / adc mcp project server

ADC MCP Project Server

Provides enterprise MCP routing across text analysis, code review, sentiment, and knowledge management using local Ollama models.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "harunraseed07-adc_mcp_project": {
      "command": "python",
      "args": [
        "-m",
        "mcp_servers.text_analysis_server"
      ]
    }
  }
}

You run an MCP server that coordinates multiple specialized AI services over a lightweight, standardized JSON-RPC protocol. It lets you route queries to the right analysis tools, run local AI models, and present results in a clean web interface for enterprise-grade text analysis, code review, sentiment analysis, and knowledge management.

How to use

You interact with the MCP server through an MCP client. Start the four local analysis servers and connect the client to them. The system will route queries to the appropriate server based on the task you request and the available tools. You can perform text analysis, code review, sentiment analysis, and knowledge management, and you’ll receive structured results suitable for integration into dashboards, reports, or knowledge bases.

How to install

Prerequisites you need before installing and running the MCP system.

# 1) Install Python (3.8+)
# 2) Install Ollama (local AI model runtime)
# 3) Install Git (for cloning the project)

Install Python and verify the installation.

pip install --upgrade pip
python --version

Clone the MCP project, install dependencies, and verify the required model is available.

git clone https://github.com/harunraseed07/ADC_MCP_Project.git
cd ADC_MCP_Project
pip install -r requirements.txt
ollama list
ollama pull llama3.2:3b

Start the Ollama service and confirm the model is ready.

ollama serve
ollama list

Start the four MCP servers locally. Each server runs a dedicated Python module as a stdio MCP server.

python -m mcp_servers.text_analysis_server
python -m mcp_servers.code_review_server
python -m mcp_servers.sentiment_analysis_server
python -m mcp_servers.knowledge_server

Optionally, run the web interface to interact with the MCP system.

python web_app/app.py

Configuration and usage notes

The system is configured to use Ollama with the llama3.2:3b model. The MCP configuration is prepared to connect local (stdio) MCP servers and a web interface on their respective ports.

Key ports used by the MCP servers and the web interface are as follows.

- Text Analysis: port 8001 - Code Review: port 8002 - Sentiment Analysis: port 8003 - Knowledge: port 8004 - Web Interface: port 5000

Security and production notes

This setup runs locally for demonstration and testing. For production, plan to add authentication, enforce HTTPS/WSS, implement rate limiting, validate inputs, and use a proper data store for persistence.

Enable debug logging during testing to diagnose issues. In a local development environment you can set environment variables to control logging levels.

Available tools

summarize_text

AI-powered text summarization tool that condenses input text into a concise summary