prompt-sage/Documentation/PROJECT_SUMMARY.md

PromptSage Project Summary

Project Overview

PromptSage is a lightweight Mac menu bar app that observes when users are writing prompts in AI tools and offers to enhance those prompts before submission, resulting in better AI outputs.

Project Structure

prompt-magic/
├── Assets/                       # Source assets
│   ├── logo.svg                  # Main app logo
│   ├── MenuBarIcon.svg           # Menu bar icon (active state)
│   └── MenuBarIconInactive.svg   # Menu bar icon (inactive state)
│
├── Documentation/                # Project documentation
│   ├── Marketing_Plan.md         # Comprehensive marketing strategy
│   └── PROJECT_SUMMARY.md        # This file
│
├── PromptSage.xcodeproj/         # Xcode project
│   └── project.pbxproj           # Project configuration
│
├── Resources/                    # App resources
│   ├── Assets.xcassets/          # Asset catalog
│   │   ├── AppIcon.appiconset/   # App icon
│   │   ├── MenuBarIcon.imageset/ # Menu bar active icon
│   │   └── MenuBarIconInactive.imageset/ # Menu bar inactive icon
│   │
│   ├── Info.plist                # App configuration
│   └── MainMenu.xib              # Main menu interface
│
├── Sources/                      # Swift source code
│   ├── AppDelegate.swift         # Main app entry point
│   ├── PromptDetector.swift      # Text monitoring and prompt detection
│   ├── PromptEnhancer.swift      # LLM integration for prompt enhancement
│   ├── PromptView.swift          # SwiftUI interface for enhancement
│   └── Preferences.swift         # Preferences window UI
│
└── README.md                     # Project README

Implementation Details

Core Components

1. AppDelegate.swift

   • Entry point for the application
   • Sets up menu bar icon and menu
   • Coordinates other components
   • Manages app lifecycle

2. PromptDetector.swift

   • Monitors text input across applications
   • Detects potential AI prompts
   • Has configurable detection patterns
   • Uses Accessibility APIs

3. PromptEnhancer.swift

   • Connects to LLM APIs (OpenAI, Anthropic, Google, Ollama)
   • Optimizes prompts based on selected style
   • Handles API authentication and rate limiting
   • Provides analytics on improvements

4. PromptView.swift

   • SwiftUI interface for the enhancement window
   • Shows before/after comparison
   • Provides metrics on prompt improvements
   • Allows manual editing of enhanced prompts

5. Preferences.swift

   • Settings interface for the app
   • LLM provider configuration
   • Detection and enhancement preferences
   • Custom patterns and tools configuration

Key Features

1. System-wide Integration

   • Works across all applications
   • Detects AI interfaces automatically
   • Accessible from menu bar

2. Intelligent Enhancement

   • Multiple enhancement styles (balanced, concise, detailed, etc.)
   • Support for various LLM providers
   • Analytics on improvement metrics

3. Privacy-Focused

   • Local processing where possible
   • Clear permissions model
   • No data retention

4. Customization

   • User-defined detection patterns
   • Configurable AI tools list
   • Adjustable enhancement settings

Technical Notes

Requirements

• macOS 13.0 (Monterey) or later
• Accessibility permissions
• Internet connection (for cloud LLMs)

APIs Used

• macOS Accessibility API for text monitoring
• OpenAI API for GPT models
• Anthropic API for Claude models
• Google AI API for Gemini models
• Ollama API for local models

Build Instructions

1. Open PromptSage.xcodeproj in Xcode
2. Configure signing identity
3. Build and run (Cmd+R)

Deployment

The app can be distributed through:

1. Mac App Store (primary)
2. Direct download from website
3. Enterprise distribution

Next Steps

1. Implement MainMenu.xib
2. Complete Accessibility API integration
3. Add unit and UI tests
4. Set up CI/CD pipeline
5. Create installer package