Quick Start
1. Start the Agent
Section titled “1. Start the Agent”katachi startOn first run, you’ll be guided through authentication:
First-time setup detected!
Please visit: https://katachi.live/deviceEnter code: ABCD-1234
Waiting for authorization...✓ Authorized!✓ Tunnel provisioned: agent-abc123.katachi.live✓ Katachi agent started successfully
PID: 12345 Port: 7676 Logs: ~/.katachi/logs2. Open the Web UI
Section titled “2. Open the Web UI”Visit katachi.live and log in with the same account you just authorized.
You’ll see your agent appear as online in the dashboard. From here you can:
- Start a new coding session
- Select an AI provider and model
- Open the file editor
- Chat with your AI assistant
3. Start Coding
Section titled “3. Start Coding”Create a new session, enter a Git repository URL, and select your AI provider. The Katachi agent will clone the repository into ~/.katachi/repos on your machine, giving the AI full access to the codebase and your development tools — all through the secure tunnel.
What’s Happening Behind the Scenes
Section titled “What’s Happening Behind the Scenes”- The agent starts a local HTTP server on port 7676
- A Cloudflare tunnel connects your machine to the Katachi backend
- The web UI sends requests through the backend to your agent
- AI responses stream back in real-time via Server-Sent Events (SSE)
Next Steps
Section titled “Next Steps”- Configure your AI providers — Set up OAuth or API keys
- Learn the CLI commands — Master the
katachicommand - Understand the architecture — See how it all fits together