mcp-data-analyst-Local_MCP_server

GurleenSingh0701/mcp-data-analyst-Local_MCP_server

3.2

If you are the rightful owner of mcp-data-analyst-Local_MCP_server and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to dayong@mcphub.com.

The Enterprise Autonomous Agent Platform is a local-first Agentic AI Platform utilizing the Model Context Protocol (MCP) to securely orchestrate LLMs for complex tasks on local infrastructure.

🚀 Enterprise Autonomous Agent Platform (MCP)

A local-first Agentic AI Platform built with the Model Context Protocol (MCP). This system orchestrates LLMs (Claude/GPT-4) to perform complex Engineering and Data Science tasks securely on local infrastructure without data egress.

Architecture Status Status Python

🧠 System Architecture

This project is not a chatbot; it is a multi-modal tool orchestration engine capable of:

1. 📊 Autonomous Data Science Analyst

  • Capability: Ingests raw CSV datasets (shopping_behavior_updated.csv) to perform schema inference and exploratory data analysis.
  • Machine Learning: Automated K-Means Clustering (Scikit-Learn) to detect customer segments (e.g., "High-Spend/Low-Loyalty") without human guidance.
  • Visualization: Auto-generates distribution histograms and scatter plots using Matplotlib.

2. 🗺️ Static Analysis & Architecture Mapping ("Cartographer")

  • Capability: Parses Python Abstract Syntax Trees (AST) to map legacy codebases.
  • Graph Theory: Uses NetworkX to build Directed Acyclic Graphs (DAGs) of module dependencies, automatically identifying "God Objects" and high-coupling risks.

3. 🛡️ Self-Healing QA Engineer

  • Capability: Autonomous Test-Driven Development (TDD) loop.
  • Workflow: Reads source code -> Generates pytest cases -> Executes tests -> Analyzes stderr -> Refactors code to fix failures.

🛠️ Tech Stack

  • Core Protocol: Model Context Protocol (MCP)
  • Runtime: Python 3.12 (managed via uv)
  • Data Intelligence: Pandas, NumPy, Scikit-Learn
  • Static Analysis: Python AST, NetworkX
  • Testing: Pytest, Subprocess

🚀 How It Works

The system runs as a local MCP Server that exposes tool primitives to an MCP Client (Claude Desktop / VS Code Cline).

Example Workflow (Data Science):

"Analyze the sales data, run a 4-cluster K-Means segmentation on 'Purchase Amount' vs 'Frequency', and visualize the results."

Example Workflow (QA Automation):

"Read the utils.py module, generate a comprehensive test suite covering edge cases, and ensure all tests pass."

📂 Project Structure

├── server.py                # Core MCP Server & Tool Definitions
├── shopping_behavior.csv    # Proprietary Dataset
├── mock_project/            # Target Legacy Codebase for Analysis
│   ├── main.py
│   ├── database.py          # "God Object" Dependency
│   ├── auth.py
│   ├── utils.py
│   ├── test_utils.py        # Autonomously generated test suites
│   └── payment.py
└── pyproject.toml           # Dependencies (uv)