# MCP Explained — an interactive guide to every MCP message on the wire

> A step-by-step, animated walk through the Model Context Protocol, with every message shown as it travels.

- Org: Runlayer
- Date: 2026-11-05
- Category: tutorial, resource
- Source: https://mcp-explained.com/
- Canonical page: https://ai-tldr.dev/releases/runlayer-mcp-explained/

## In short

MCP Explained is a free, animated guide from Runlayer that steps through the Model Context Protocol one JSON-RPC message at a time, from lifecycle and transports to auth, tools and known attacks. It hit 279 points on HN.

## What it is

MCP Explained is an interactive explainer of the Model Context Protocol (MCP), the open standard that connects AI apps to tools and data. Its lessons cover JSON-RPC, the session lifecycle, transports, authorization, tools, resources, prompts, sampling, elicitation, roots, progress and tasks, and each lesson links to the official spec.

## How it works

Each lesson plays a message flow between the host (the AI app), the MCP server and the model, showing calls like tools/list and tools/call step by step. A 'spec time machine' replays each flow as it worked in five spec versions, from 2024-11-05 to 2026-07-28, and a 'Break it' section demonstrates tool poisoning, tool shadowing, token passthrough, prompt injection and elicitation phishing.

## Why it matters

Developers building or securing agents can see the actual wire traffic instead of reading the spec cold, including how attacks on MCP servers work. HN commenters were split: some praised the clear opening, while others found the multi-pane lesson view dense for beginners.

## Who it's for

developers building or securing MCP servers and agents

## Try it

https://mcp-explained.com/

## Links

- [MCP Explained](https://mcp-explained.com/)
- [Source Code](https://github.com/Runlayer/mcp-explained)
- [HN Discussion](https://news.ycombinator.com/item?id=50379249)