MCP (Model Context Protocol) Server

What is MCP (Model Context Protocol) Server?

This server provides a unified interface for interacting with multiple services including GitHub, Azure Log Analytics, Google, Azure OpenAI, Azure Vision, and PostgreSQL. It is designed for developers who need a single MCP endpoint to access diverse APIs and data sources.

How to use MCP (Model Context Protocol) Server?

Install dependencies via pip install -r requirements.txt, create a .env file with required tokens (GitHub PAT, Azure credentials, etc.), then run python server.py. The server starts on http://0.0.0.0:6277 by default. Docker is also supported: build with docker build -t mcp-server . and run with docker run -p 6277:6277 --env-file .env mcp-server.

Key features of MCP (Model Context Protocol) Server

• Retrieve files and directory listings from GitHub repositories
• Monitor GitHub workflow runs and code search
• Run Kusto Query Language (KQL) queries on Azure Log Analytics
• Read PostgreSQL databases with read-only user support
• Search Google and YouTube, get YouTube video transcripts
• Analyze images via Azure Vision and generate responses via Azure OpenAI

Use cases of MCP (Model Context Protocol) Server

• Search code across GitHub repositories and inspect commit diffs
• Query Azure Log Analytics for operational insights or security logs
• Get structured content from websites in Markdown format
• Retrieve YouTube video transcripts for analysis or summarization
• Execute read-only queries on PostgreSQL databases

FAQ from MCP (Model Context Protocol) Server

What prerequisites are needed?

Python 3.8+, optional Docker, and service-specific tokens: GitHub PAT (repo scope), Azure AD client ID/secret/tenant, Azure OpenAI key/endpoint, Azure Vision key/endpoint, and PostgreSQL credentials.

How do I set up tokens?

Create a .env file with the required variables as described in the README (e.g., GITHUB_TOKEN, AZURE_CLIENT_ID, etc.). Each service’s token setup is detailed in the respective section.

Can I run the server in Docker?

Yes. Build the Docker image with docker build -t mcp-server . and run it with docker run -p 6277:6277 --env-file .env mcp-server. Ensure your .env file is present.

Does the PostgreSQL access allow write operations?

No. The recommended practice is to use a database user with SELECT-only privileges. The read_db tool is designed for read-only queries.

What transports or authentication does the server use?

The server listens on HTTP (0.0.0.0:6277). Authentication for each external service is handled via tokens/credentials stored in the .env file; no additional transport-level security is described.