Skip to main content

Onboarding

GenieBuilder includes a guided onboarding wizard that helps you get started quickly. The setup wizard launches automatically on first run and walks you through essential configuration steps.

Onboarding Demo​

Watch this demo video to see the onboarding wizard in action:

Setup Wizard Flow​

The onboarding experience consists of 6 steps:

Step 1: Welcome and Project Selection​

The welcome screen introduces GenieBuilder and lets you select a project workspace.

Onboarding Step 1 - Welcome

What you'll see:

  • Overview of GenieBuilder features
  • Introduction to Agentic Software Engineering
  • Project selection options

Options:

  • Open Project: Browse and select a local project folder
  • Skip for now: Continue without a project (you can set this later in Settings)

Your action: Click Get Started to begin configuration.

Tip: Selecting a project now lets you immediately start exploring files and running agentic workflows.

Step 2: Hardware Check​

The wizard checks your system capabilities for running local AI models.

Onboarding Step 2 - Hardware Check

What's checked:

  • Available memory (RAM) for model loading
  • Graphics acceleration (GPU) availability
  • Free disk space for model downloads

Results:

  • Green indicators show capabilities that meet recommended thresholds
  • Yellow indicators show areas that may limit performance
  • You'll get recommendations for the best local AI model experience

Step 3: Local LLM Models​

Configure local AI options for offline or privacy-focused usage.

Onboarding Step 3 - Local Models

Local options:

  • Ollama - Run AI models entirely on your computer (recommended)
  • LM Studio - GUI-based local model management
  • llama.cpp - Direct integration with custom builds

Setup:

  • If you already have Ollama installed, GenieBuilder will detect it automatically
  • If not, you can install it later and configure it in Settings
  • Select which model to use from your locally available models

Benefits of local models:

  • Complete privacy - your code stays on your machine
  • Works offline without internet connection
  • No API costs

Step 4: Cloud LLM Providers​

Configure hosted AI providers for immediate use.

Onboarding Step 4 - Cloud Providers

Supported providers:

  • OpenAI - GPT-4 and other GPT models
  • Anthropic - Claude models
  • Google - Gemini models

What you need:

  • An API key from your chosen provider(s)
  • API keys are stored securely using your system's secure storage

Tip: You can skip this step and configure cloud providers later in Settings. You only need at least one provider (cloud or local) to start using GenieBuilder.

Step 5: CLI Runners​

Configure CLI-based AI agents for specialized agentic workflows.

Onboarding Step 5 - CLI Runners

What are CLI Runners? CLI Runners let you integrate powerful command-line AI tools like:

  • Claude Code - Anthropic's agentic coding assistant
  • Kimi CLI - Fast, efficient code generation
  • Codex - OpenAI's code review tool

Setup:

  • CLI Runners are optional but recommended for advanced agentic workflows
  • If you have them installed, GenieBuilder will detect them automatically
  • You can add or configure them later in Settings

Step 6: Complete​

Review your configuration and finish setup.

Onboarding Step 6 - Complete

Summary shown:

  • Project path (if selected)
  • Configured cloud providers
  • Configured local providers
  • MCP integration status

Your action: Click Launch to enter the main application.

After Setup​

Once you complete the wizard:

  1. Main workspace opens with your selected project (if any)
  2. Your settings are saved and will be remembered next time
  3. You can start using multi-tab AI chat, agentic workflows, and task management

Changing Settings Later​

You can modify any configuration after onboarding:

  • Click the Settings button (gear icon) in the header
  • Add or remove AI providers
  • Configure MCP integrations
  • Adjust appearance and behavior preferences

Getting Help​

If you encounter issues during setup:

Troubleshooting​

Wizard Stuck on Loading​

If the wizard hangs during setup:

  1. Check your internet connection (for cloud provider validation)
  2. If testing local providers, make sure Ollama is running
  3. Try restarting the application

Project Selection Fails​

If selecting a project folder doesn't work:

  1. Make sure you have permission to access the folder
  2. Try selecting a different folder
  3. You can skip this step and set the project later

Provider Connection Issues​

If connecting to AI providers fails:

  1. Double-check your API key is correct
  2. For local providers, verify the service is running (Ollama, etc.)
  3. You can skip and configure providers later in Settings

MCP Connection Issues​

If tool integrations fail to connect:

  1. You can skip this step and configure MCP later
  2. Check the MCP settings in the Settings panel after setup
  3. Some MCP features are optional for basic usage