Skip to main content

Quick Start

Get up and running with GenieBuilder in minutes.

1. Complete Onboarding​

When you first launch GenieBuilder, the setup wizard will guide you through:

  • Selecting a project workspace
  • Configuring cloud AI providers (OpenAI, Anthropic, Gemini)
  • Setting up local AI (Ollama) if desired
  • Configuring MCP tools

You can skip any step and configure it later in Settings.

2. Open a Workspace​

Select a project folder using the file explorer. GenieBuilder will:

  • Show a .gitignore-aware file tree
  • Allow you to open files in the built-in editor
  • Track tasks in the right panel (if Backlog is configured)

3. Configure an AI Provider​

Choose from cloud or local providers:

Cloud Providers:

  • OpenAI — GPT-5.4 and other GPT models
  • Anthropic — Claude 4.6 Opus or Sonnet
  • Google — Gemini 3.1 Pro and Flash

Local Providers:

  • Ollama — Run models locally for privacy
  • LM Studio — GUI-based local model management
  • llama.cpp — Direct integration with custom builds

Add your API keys in Settings → Providers. Keys are stored securely using macOS Keychain.

4. Start a Chat​

  1. Open the chat panel (bottom of screen)
  2. Click New Tab to start a conversation
  3. Select your provider and model
  4. Type your prompt and press ⌘+Enter to send

Each chat tab can use a different provider, letting you compare responses or use specialized models for different tasks.

5. Try a Workflow​

  1. Click the Workflow button in the header
  2. Select an example from the sidebar (try "One Prompt" first)
  3. Click Run to execute
  4. Watch the workflow progress in real-time

Workflows can automate tasks like code review, bug fixes, or documentation generation.

Next Steps​