MCP (Model Context Protocol) Server

A MCP server that provides a unified interface for interacting with multiple services, currently supporting GitHub, Azure Log Analytics, Google, Azure OpenAI, Azure Vision, PostgreSQL, and more.

🔧 Features

GitHub Integration

• File/Folder Access: Retrieve files or directory listings from GitHub repositories
• Workflow Monitoring: Fetch recent workflow runs and their statuses
• Code Search: Search for code across repositories
• Repository Structure: Get the complete file structure of a repository
• Commit History: View commit history with optional file filtering
• Commit Diffs: Inspect file-level changes for specific commits

Azure Integration

• Log Analytics Querying: Run Kusto Query Language (KQL) queries against Azure Log Analytics workspaces
• Secure Authentication: OAuth2 client credentials flow for secure API access

Utilities

• Timestamp Generation: Get current UTC timestamp with microsecond precision
• Website Content: Get the content of a website in Markdown format

Database Integration

• Database Reading: Read a database from a PostgreSQL server

⚠️ Use a DB user with SELECT-only privileges

Google Integration

• Google Search: Search Google for a query
• YouTube Search: Search YouTube for a query
• YouTube Transcript: Get the transcript of a YouTube video

Azure Vision Integration

• Image Analysis: Get the analysis of an image

LLM Integration

• Azure OpenAI: Get a response from Azure OpenAI

📦 Prerequisites

• Python 3.8+
• Docker (optional)
• Tokens & credentials:
  • GitHub PAT (repo scope)
  • Azure AD App: client ID, secret, tenant ID
  • Azure OpenAI Key & Endpoint
  • Azure Vision Key & Endpoint
  • PostgreSQL credentials

Installation

1. Clone the repository:

   git clone https://github.com/DrFloflo/MCP-github.git
   

2. Install dependencies:

   pip install -r requirements.txt
   

3. Create a .env file in the project root with the following variables:

   # GitHub Configuration
   GITHUB_TOKEN=your_github_token
   
   # Azure Configuration
   AZURE_CLIENT_ID=your_client_id
   AZURE_CLIENT_SECRET=your_client_secret
   AZURE_TENANT_ID=your_tenant_id
   
   # Azure OpenAI Configuration
   AZURE_OPENAI_ENDPOINT=your_azure_openai_endpoint
   AZURE_OPENAI_KEY=your_azure_openai_key
   AZURE_OPENAI_API_VERSION=your_azure_openai_api_version
   
   # Azure Vision Configuration
   VISION_ENDPOINT=your_vision_endpoint
   VISION_KEY=your_vision_key
   
   # Database Configuration
   POSTGRES_HOST=your_postgres_host
   POSTGRES_PORT=your_postgres_port
   POSTGRES_DB=your_postgres_db
   POSTGRES_USER=your_postgres_user
   POSTGRES_PASSWORD=your_postgres_password
   
   

🔐 Token Setup

GitHub

1. Go to GitHub Personal Access Tokens
2. Generate a new token with the repo scope
3. Add this to your .env file:

Azure AD (Log Analytics)

1. Register a new app in Azure Active Directory
2. Copy the following:
   • Tenant ID (Directory ID)
   • Client ID (Application ID)
   • Create and copy a new client secret (under Certificates & secrets)
3. Add to .env:

   AZURE_TENANT_ID=your_tenant_id
   AZURE_CLIENT_ID=your_client_id
   AZURE_CLIENT_SECRET=your_client_secret
   

Log Analytics Workspace

1. Go to your Log Analytics Workspace in Azure
2. Copy your Workspace ID
3. Ensure your Azure AD Application has access:
   • Go to your app in Azure AD
   • Navigate to API Permissions
   • Click Add a permission
   • Choose Microsoft Graph > Delegated permissions
   • Select User.Read and confirm

Azure OpenAI

1. Go to your Azure OpenAI resource (or Foundry)
2. Copy the endpoint and key
3. Add this to .env:

   AZURE_OPENAI_ENDPOINT=your_azure_openai_endpoint
   AZURE_OPENAI_KEY=your_azure_openai_key
   AZURE_OPENAI_API_VERSION=your_azure_openai_api_version
   

Azure Vision

1. Go to your Azure Vision resource (or Vision)
2. Copy the endpoint and key
3. Add this to .env:

   VISION_ENDPOINT=your_vision_endpoint
   VISION_KEY=your_vision_key
   

PostgreSQL

1. Go to your PostgreSQL server
2. Copy the host, port, database, user, and password
3. Add this to .env:

   POSTGRES_HOST=your_postgres_host
   POSTGRES_PORT=your_postgres_port
   POSTGRES_DB=your_postgres_db
   POSTGRES_USER=your_postgres_user
   POSTGRES_PASSWORD=your_postgres_password
   

Usage

Running the Server

python server.py

The server will start on http://0.0.0.0:6277 by default.

Docker

Build the Docker image:

docker build -t mcp-server .

Run the container:

docker run -p 6277:6277 --env-file .env mcp-server

🧰 Available API Tools

GitHub Tools

• get_github_file_folder: Get file contents or list directories
• get_workflow_runs: Retrieve recent workflow runs
• search_codebase: Search code across repositories
• get_file_structure: Get complete repository structure
• get_commit_history: View commit history
• get_commit_diff: View file-level changes in a commit

Azure Tools

• run_log_analytics_query: Execute KQL queries against Azure Log Analytics

PostgreSQL Tools

• read_db: Query data from PostgreSQL

⚠️ Use a database user with read-only (SELECT) access

Google Tools

• search_google: Search Google for any query
• search_youtube: Search YouTube videos
• get_youtube_transcript: Get transcript from a YouTube video

Azure Vision Tools

• get_image_analysis: Analyze image content using Azure Vision

Azure OpenAI Tools

• get_azure_openai_response: Generate response using Azure OpenAI (LLM)

Utility Tools

• get_current_utc_timestamp: Get the current UTC timestamp
• get_website_content: Fetch and convert website HTML to Markdown

Code Tools

• execute_python_code: Execute a Python code snippet securely