Skip to main content

1. Install ADE

Download the DMG, drag ADE into /Applications, and launch it.

2. Connect a provider

Add an Anthropic or OpenAI-compatible key so ADE can run agents.

3. Open your repo

Point ADE at a Git repository and let onboarding create the local ADE state.

4. Create a lane

Fork a task into its own isolated worktree.

5. Run your first agent

Use Agent Chat for a fast task or launch a Worker run for structured execution.

6. Learn the model

Read the concepts page once so the rest of the docs make sense.
1

Install the app

ADE currently ships a macOS beta installer. Download the DMG from GitHub Releases and drag ADE.app into /Applications.
2

Add one AI provider

Open Settings and save at least one provider credential. Anthropic is the fastest path to a working setup.
3

Open a Git repository

Use Open Repository on the welcome screen or press Cmd+O.
4

Create a lane

Create a new lane from your base branch so the first task does not run in the primary checkout.
5

Run a small task

Start with one contained change, like adding a health endpoint or updating a small component. That gives you a clean first pass through the chat, file diff, and commit flow.
Small, easy to review, and exercises the full loop: open a lane → run an agent → inspect the diff → commit.

Next Steps

Key Concepts

Understand every core concept in ADE: Lanes, Packs, Workers, the CTO, Automations, and more.

Lanes

Deep dive into lane types, templates, environment isolation, port allocation, and proxy routing.

Workers

Learn how to configure phase profiles, set budget limits, handle interventions, and export audit bundles.

CTO Agent

Set up the always-on CTO agent with Linear sync and worker agent delegation.

Automations

Create trigger-driven automation rules that fire agents in response to git events, Linear updates, and schedules.

Configuration

Configure AI providers, ADE CLI, permissions, and the full ade.yaml schema.