# Yertle

> Yertle is a collaborative visual workspace for software architecture. It gives engineering teams — and the AI agents working alongside them — a version-controlled, machine-readable map of how their systems actually fit together. Think Git for system design: immutable commits, branches, and pull requests over a graph of services, connections, and documentation, accessible through a web app, CLI, and MCP server.

## Core concepts

- [System Context for AI Agents](https://yertle.com/system-context-for-agents): Why AI coding agents need structured system context beyond the code they can read, and what that context looks like.
- [Agentic System Design](https://yertle.com/agentic-system-design): How to design and document architecture so AI agents can reason about it, not just humans.
- [Version Controlled Architecture](https://yertle.com/version-controlled-architecture): Git-like branching, commits, and pull requests applied to system design instead of source code.
- [Context Graph](https://yertle.com/context-graph): The graph model Yertle uses to represent nodes, connections, and hierarchy across a software system.
- [Architecture MCP Server](https://yertle.com/architecture-mcp-server): An MCP server that exposes an organization's architecture to AI agents as queryable, structured context.

## Product

- [Home](https://yertle.com/): Overview of Yertle — the visual workspace and context graph for software architecture.
- [Early Access](https://yertle.com/early-access): Request access to Yertle.

## Blog

- [Blog index](https://yertle.com/blog)
- [Your Architecture Diagrams Are Lying to You](https://yertle.com/blog/your-architecture-diagrams-are-lying-to-you): Why Confluence diagrams and tribal knowledge decay, and how version-controlled system context fixes it.
- [Your AI Agent Can Write Code. It Can't Read Your Architecture.](https://yertle.com/blog/your-ai-agent-cant-read-your-architecture): AI coding agents lack structured awareness of how systems connect — stale wikis don't close the gap.

## Developer resources

- [Documentation](https://docs.yertle.com): API, CLI, and MCP reference.
- [CLI (Homebrew)](https://docs.yertle.com): `brew tap model-context/yertle && brew install yertle` — scriptable access to orgs, nodes, trees, and an ASCII canvas renderer.
- [MCP server](https://yertle.com/architecture-mcp-server): Connect AI agents to your Yertle organization via the Model Context Protocol.