Quack is a native desktop GUI for Claude Code that lets you work on multiple AI coding sessions simultaneously. It's built with Tauri + React and provides a visual interface for managing Claude Code projects, agents, terminals, and more.

Yes! Quack is completely free — unlimited agents, no limits, no catches. We offer personal setup sessions (€79), an Expert program with resell rights (€149), custom skill packs, and enterprise consulting for teams (from €299). The app itself is free forever.

How many Claude Code sessions can I run?

There's no limit — Quack is free with unlimited agents. Run 10, 20, or 30+ Claude Code sessions in parallel. The only limit is your machine's resources.

Does Quack work with other AI tools?

Quack is built exclusively for Claude Code. It's an opinionated choice — we believe building deeply for one ecosystem is better than building shallowly for many. That said, Quack works perfectly alongside your IDE (VS Code, Cursor, etc.).

What platforms does Quack support?

Quack is available for macOS 12.0 or later (Apple Silicon and Intel) and Windows x64. Linux is planned for a future release.

Yes. All your data stays on your machine. Quack uses local storage only — no cloud sync, no external servers, no tracking. Your code, conversations, and projects never leave your computer.

Can I make money with Quack?

Yes! With the Quack Expert plan (€149), you get 2 hours of advanced setup with the founder, a dedicated Discord channel, your profile on our Experts page, and the right to resell Quack consulting to your own clients at your own rate. One client covers the cost — everything after that is profit.

Quack is free, so there's nothing to refund for the app itself. For Gumroad purchases (skill packs, personal setup), we offer a 30-day money-back guarantee — just email us at quack@quack.build, no questions asked.

Ai Image Generation

AI Image Generation brings visual creativity directly into your workflow. With the openai image gen skill, agents can generate custom images, illustrations, and graphics using OpenAI's DALL-E model - perfect for landing pages, marketing materials, documentation, or any project that needs visual assets.

Overview

When an agent has access to the image generation skill, it can:

Generate images from text descriptions - Natural language prompts
Preview inline - Images appear directly in chat stream
Save automatically - Files stored in project assets folder
Open full-size - One-click viewer for detailed inspection
Work autonomously - Agent handles creative decisions

Setup

Requirements

To use AI Image Generation, you need:

OpenAI API Key - From platform.openai.com
Quack Settings Configuration - Add key to settings
Image Generation Skill - Assigned to your agent

Getting an OpenAI API Key

Go to platform.openai.com/api-keys
Click Create new secret key
Name it (e.g., "Quack Image Generation")
Copy the key (starts with sk-...)

Configuring Quack

Open Quack Settings (Cmd+, or Ctrl+,)
Navigate to AI Assistant section
Find OpenAI API Key field
Paste your API key
Click Save

Your API key is:

Stored locally - Never sent to Quack servers
Encrypted on macOS - Uses system keychain
Used only for image generation - Not shared with other services

Assigning the Skill

The openai image gen skill needs to be assigned to your agent:

Option 1: Via Agent Settings

Open the Agent Context panel
Click Edit Agent
Go to Skills section
Add openai image gen skill
Save

Option 2: Via Rules Add to agent's CLAUDE.md or rules:

Use the openai image gen skill for visual assets

Generating Images

Basic Usage

Simply ask the agent to create an image:

You: "Create a hero image for our landing page"

Agent: "I'll generate a hero image for your landing page."

[Generating image...]

[Image preview appears inline]

Image saved to: /project/assets/hero-landing-page.png

Providing Details

More specific prompts produce better results:

You: "Generate a minimalist illustration of a rocket launching,
      with a gradient background from purple to blue,
      suitable for a SaaS landing page"

Agent: "Generating illustration..."

[Image appears with specified style]

Multiple Images

Request variations or different assets:

You: "Create 3 different icon styles for our app:
      1. Modern and minimalist
      2. Playful and colorful
      3. Professional and clean"

Agent: "I'll generate three icon variations..."

[Three images appear in sequence]

Image Preview

Inline Preview

Generated images appear directly in the chat stream:

┌─────────────────────────────────────────────────────┐
│  Agent: "Here's the hero image:"                    │
│                                                      │
│  ┌─────────────────────────────────────┐           │
│  │                                      │           │
│  │    [Generated Image Preview]         │           │
│  │      600x400px thumbnail             │           │
│  │                                      │           │
│  └─────────────────────────────────────┘           │
│                                                      │
│  [Open in tab] hero-landing-page.png                │
└─────────────────────────────────────────────────────┘

Full-Size Viewer

Click Open in tab to view the image at full resolution:

New tab opens with image viewer
Zoom in/out with mouse wheel
Pan around the image
See full metadata (dimensions, prompt, model)

Image Format

Generated images are:

Format: PNG (lossless, transparent backgrounds supported)
Size: 1024x1024px (default) or custom dimensions
Quality: High-resolution, production-ready

File Management

Automatic Saving

Images are automatically saved to your project:

Default location:

/your-project/assets/images/

File naming:

Based on description or content
Sanitized for filesystem compatibility
Numbered if duplicates exist

Example structure:

project/
├── assets/
│   └── images/
│       ├── hero-landing-page.png
│       ├── feature-icon-dashboard.png
│       ├── team-photo-placeholder.png
│       └── background-gradient-blue.png

Custom Paths

Specify where to save:

You: "Generate a logo and save it to /public/logo.png"

Agent: "Generating logo and saving to /public/logo.png..."

Organizing Assets

Create organized asset folders:

You: "Generate illustration assets for the features section.
      Save them in /assets/features/"

Agent: "I'll create illustrations for each feature..."

Result:
/assets/features/
├── analytics-chart.png
├── collaboration-team.png
└── security-shield.png

Use Cases

Landing Pages

Prompt:

"Create a modern, vibrant hero image for a productivity app landing page.
 Include elements like calendars, checkboxes, and notifications.
 Use a gradient from #FF6B35 to #004E89."

Result: Custom hero image matching your brand colors.

Documentation

Prompt:

"Generate a simple diagram showing data flow:
 User → API → Database → Response.
 Use arrows and boxes, keep it minimal."

Result: Clean visual aid for your docs.

Placeholder Content

Prompt:

"Create 5 diverse team member placeholder photos -
 professional headshots with neutral backgrounds"

Result: Placeholder images for development.

Marketing Materials

Prompt:

"Design a social media banner (1200x630) showcasing our new feature.
 Include text space on the left, visual element on the right.
 Professional but friendly style."

Result: Social media ready graphics.

App Icons and Assets

Prompt:

"Generate a set of feature icons for our dashboard:
 - Analytics (chart icon)
 - Settings (gear icon)
 - Users (people icon)
 - Reports (document icon)
 Keep them consistent in style, use single color #004E89"

Result: Cohesive icon set.

Advanced Features

Style Consistency

Maintain visual consistency across images:

You: "Use the same style as the previous image but for a different scene"

Agent: *Generates new image matching previous aesthetic*

Refine images through conversation:

You: "Make the background lighter"
You: "Add more contrast"
You: "Perfect! Save this version"

Batch Generation

Generate multiple assets efficiently:

You: "Generate all the images needed for the landing page:
      - Hero section background
      - 3 feature illustrations
      - Footer decoration
      - About team photo placeholder"

Agent: *Creates all 6 images systematically*

Integration with Website Building

When building sites, agents automatically generate needed visuals:

You: "Build a landing page for a fitness app"

Agent:
1. Creates HTML structure
2. Generates hero image with fitness theme
3. Creates feature section icons
4. Generates testimonial avatars
5. Saves all assets to /public/images/
6. Updates HTML to reference generated images

Best Practices

Writing Effective Prompts

Be specific about:

Style: "minimalist", "photorealistic", "illustrated", "abstract"
Colors: Hex codes, color names, or "match brand colors"
Mood: "professional", "playful", "serious", "energetic"
Elements: What should be in the image
Purpose: Where the image will be used

Good prompt:

"Create a minimalist illustration of a dashboard interface,
 using flat design style with #FF6B35 accent color,
 for a SaaS product hero section"

Vague prompt:

"Make a picture for the website"

When to Generate Images

Good use cases:

Prototyping and mockups
Placeholder content during development
Custom illustrations for unique concepts
Marketing materials
Social media graphics

When NOT to use:

Photos of real people you know (use real photos)
Product photography (use actual product photos)
Logos requiring legal trademark protection
Content requiring precision accuracy (technical diagrams)

Cost Awareness

Image generation uses OpenAI credits:

Standard: ~$0.02 per image (1024x1024)
HD Quality: ~$0.08 per image
Monitor usage in OpenAI dashboard
Set spending limits if needed

File Organization

Create a clear structure:

project/
├── public/
│   ├── images/          # Public-facing images
│   └── icons/           # UI icons
├── assets/
│   ├── generated/       # AI-generated assets
│   ├── originals/       # Source files
│   └── optimized/       # Compressed versions