Mastering Model Context Protocol (MCP)
From Your First Python Server to Production
by Kapil Dev V
From a Python function to a Model Context Protocol server that a real AI app calls correctly, then on to one you test, ship and run behind a gateway for a whole team. It teaches the current, stateless revision of the specification, dated 28 July 2026, through one server that grows a chapter at a time.
The Kindle eBook. Amazon.in in India, Amazon.com outside India.
Inside the book
- Host, client and server, and which one you are actually writing
- JSON-RPC, HTTP and the stateless July 2026 revision, read off the wire
- How type hints and docstrings become a schema a model calls correctly
- Resources and prompts, and the cost of making everything a tool
- Tests, the Inspector, and reading the log when a call fails
- Plugging the server into Claude Code, Claude Desktop, VS Code and Cursor
- Streamable HTTP with a token, and the pre-exposure checklist
- The gateway pattern: many servers behind one URL
- A wheel, a dry-run publish and a container with everything inside
- A capstone: a second server, designed and shipped from a one-line brief
For anyone who can write a Python function and has used an AI assistant, but has never built a tool for one. It uses Python 3.12 and fastmcp 4.1.0.
