Getting Started

From zero to a running AI chat app in 60 seconds. Three commands, zero boilerplate.

Prerequisites

Flutter 3.7+ and Dart 3.7+
A Flutter project (flutter create my_app)

Quick Start

1. Install the CLI

$ dart pub global activate flai_cli

2. Initialize and configure

Interactive prompts set your app name, assistant name, and theme preset.

$ flai init

You'll be asked:

App name — Your app's display name (default: FlAI Chat)
Assistant name — The AI assistant's name (default: Assistant)
Theme — Choose from dark, light, ios, or premium

These values are stored in flai.yaml and automatically applied to all generated code.

3. Generate the complete app

One command installs 7 bricks (83 source files) and generates a ready-to-run main.dart.

$ flai add app_scaffold

This installs:

auth_flow — Login, register, forgot password, verification, reset (6 screens)
onboarding_flow — Splash, naming, multi-select pills, reveal animation
chat_experience — Composer with voice, model selector, attachments
sidebar_nav — Drawer, conversation list with grouping, settings
message_bubble — Markdown rendering, code copy, thinking blocks
typing_indicator — Animated loading dots
app_scaffold — GoRouter wiring, session persistence, home controller

4. Run

$ flutter pub get && flutter run

Your app is running with mock auth. No main.dart to write — it was generated for you with your chosen theme and branding.

Connect a Backend

When you're ready for production, connect to a real AI backend:

$ flai connect cmmd

This automatically rewrites your main.dart to swap mock providers with real ones — authentication, AI streaming, conversation persistence, and voice. Zero manual wiring.

Non-Interactive Mode

For CI/CD or scripting, skip the prompts:

flai init --no-interactive --app-name "My App" --assistant-name "Nova" --theme ios
flai add app_scaffold

What's in the Box

After flai add app_scaffold, your app includes:

Auth flow with email, Apple, and Google sign-in
Configurable onboarding with splash and reveal animations
Chat with markdown rendering, code blocks with copy, and thinking panels
Sidebar with conversation history, grouping, search, and settings
Share conversation via system share sheet
Animated send button (mic/arrow morph)
Attachment menu (camera, photos, files)
Voice input (speech-to-text)
Model selector (switch AI models per-conversation)
Session persistence (stay logged in across restarts)
Scroll-to-bottom FAB and regenerate response

What's Next?

Installation — Detailed CLI options and manual setup
Theming — Customize colors, typography, and spacing
App Scaffold — Full architecture reference
Providers — Connect to OpenAI or Anthropic

Next Installation →