t128n / promptlab

⚡ A pipeline-based prompt engineering tool that helps you build LLM prompts with a modern UI.

PromptLab ⚡

PromptLab is a powerful prompt engineering and management tool built with React and TypeScript. It allows you to create, test, and manage AI prompts using a pipeline-based approach.

Features

• 🔄 Pipeline-based prompt processing
• 📝 Rich template management
• 🧩 Modular addon system
• 💾 Real-time preview
• 🎨 Modern UI with Tailwind CSS and shadcn/ui

Getting Started

1. Clone the repository:

git clone https://github.com/yourusername/promptlab.git
cd promptlab

2. Install dependencies:

bun install

3. Start the development server:

bun run dev

Project Structure

• src/ - Source code
  • lib/ - Core libraries and utilities
    • pltl/ - Prompt Lab Template Language implementation
      • templates/ - Template storage
        • prompts/ - Base prompt templates
        • assistants/ - Assistant templates
        • addons/ - Pipeline addon templates
      • interpreter.ts - PLTL interpreter
      • loader.ts - Template loader
      • types.ts - Type definitions
    • context.tsx - React context providers
    • utils.ts - Utility functions
  • components/ - React components
    • ui/ - Reusable UI components
    • app-sidebar.tsx - Sidebar component
    • copyarea.tsx - Copy text area
    • pipeline-processor.tsx - Pipeline processor
  • hooks/ - Custom React hooks
    • pltl.ts - Pipeline operation hooks
    • use-mobile.tsx - Mobile detection hook
  • App.tsx - Main application component
  • main.tsx - Application entry point
  • main.css - Global styles

The structure follows a modular approach where:

• src/lib/pltl handles core PLTL functionality
• src/components contains all React components
• src/hooks provides custom hooks for state management

Usage

1. Select a base prompt from the sidebar
2. Configure any required arguments
3. Add optional addons to modify the output
4. Enter your input text
5. View and copy the processed output

Pipeline System

The pipeline system allows you to:

• Set a base prompt template
• Add multiple addons
• Configure arguments
• Process input text through the entire pipeline

Example usage with the usePipelineOperations hook:

const { addAddon, removeAddon, replaceBasePLTL, updateArguments } = usePipelineOperations();

// Set base prompt
replaceBasePLTL(template);

// Add addon
addAddon(addon);

// Update arguments
updateArguments({ key: "value" });

Technologies

• React 18
• TypeScript
• Vite
• Tailwind CSS
• Radix UI
• Biome (Formatting)

License

MIT

Contributing

1. Fork the repository
2. Create your feature branch
3. Commit your changes
4. Push to the branch
5. Open a Pull Request

---

Built with ⚡ by Torben Haack