Agentic Tool

Github based Anthropic Model contex Protocol (MCP) server implementation for managing Agentic tool execution across multiple service integrations.

Overview

This project provides a complete MCP server implementation for integrating multiple third-party services (GitHub, Linear, etc.) into a unified tool execution platform. The server orchestrates API requests across service boundaries, enabling seamless multi-service workflows through a single interface.

Features

• Multi-Service Integration: Connect to GitHub, Linear, and other services through a unified API
• Tool Orchestration: Route tool requests to the appropriate service automatically
• Flexible Configuration: Easy service registration and configuration
• Robust Error Handling: Consistent error responses across all services

Architecture

The MCP server follows a modular architecture:

• MCP Registry: Core component that manages service and tool registration
• Service Implementations: Adapters for each integrated service (GitHub, Linear)
• Client API: Interface for executing tools across services

Getting Started

Prerequisites

• Python 3.8 or higher
• pip (Python package manager)

Installation

1. Clone this repository:

   git clone <repository-url>
   cd mcp_server_agentic_tool
   

2. Install dependencies:

   pip install -r requirements.txt
   

3. Configure your environment:

   cp .env.example .env
   # Edit .env with your service API keys
   

Running the Server

Start the MCP server:

python src/index.py

The server will start on port 5000 (configurable via PORT environment variable).

Usage

API Endpoints

• GET /services: List all available services
• GET /tools: List all available tools across all services
• POST /execute: Execute a tool by name
• POST /github/{tool_name}: Execute a GitHub tool
• POST /linear/{tool_name}: Execute a Linear tool

Example Request

Execute a GitHub tool to list issues:

curl -X POST http://localhost:5000/github/list_issues \
  -H "Content-Type: application/json" \
  -d '{}'

Client Library

The project includes a Python client library for interacting with the MCP server:

from src.client.api import client

# List all tools
tools = client.list_tools()

# Execute a GitHub tool
response = client.execute_github_tool('list_issues', {})

Integration Showcase

The project includes a showcase of how to integrate multiple services through the MCP server:

python src/client/integration_showcase.py

This demonstrates how the MCP server can be used to create seamless workflows across multiple services, enabling unified project management.

Project Structure

• src/server: Core MCP server implementation
• src/services: Service-specific implementations
• src/client: Client API and demo scripts
• src/utils: Shared utilities and type definitions