A comprehensive AI context management system with multiple interface options.
- Quick Start
- Repository Structure
- Components
- Development Workflow
- Publishing
- Integration Examples
- Contributing
- License
# 1. Build the core application
cd core/
go build -o pkt main.go
# 2. Initialize your prompt library
./pkt --init
# 3. Start using CLI commands
./pkt list # List all prompts
./pkt search "AI" # Search prompts
./pkt show prompt-id # Display specific prompt
# 1. Build and launch TUI
cd core/
go run main.go --tui
# 2. Navigate with keyboard shortcuts
# ↑/↓ or k/j - Navigate
# Enter - Open prompt
# / - Search
# Ctrl+B - Boolean search
# q - Quit
# 1. Start the API server
cd core/
go run main.go --url-server
# 2. Test the API
curl "http://localhost:8080/api/v1/health"
curl "http://localhost:8080/api/v1/prompts"
# 3. View API documentation
open "http://localhost:8080/api/docs"
# 1. Install dependencies
cd raycast-extension/
bun install
# 2. Start API server (in another terminal)
cd ../core/ && go run main.go --url-server
# 3. Import extension to Raycast
bun run dev
# 4. Launch Raycast → "Search Prompts" → Start searching!
pocket-prompt-suite/
├── core/ # Main Pocket Prompt application (Go)
│ ├── README.md # Core application documentation
│ ├── main.go # Entry point
│ ├── internal/ # Go modules and packages
│ ├── prompts/ # Default prompt library
│ ├── templates/ # Prompt templates
│ └── tests/ # Test suite
│
└── raycast-extension/ # Raycast integration (TypeScript)
├── README.md # Extension-specific documentation
├── src/ # TypeScript source code
├── assets/ # Extension assets
└── package.json # Node.js dependencies
The main Pocket Prompt application written in Go that provides:
- CLI Interface: Terminal-based prompt management
- TUI Interface: Interactive terminal UI with search and management
- HTTP Server: Web API for external integrations
- Library Management: Organize prompts, templates, and saved searches
- Git Sync: Synchronize prompt libraries across devices
A powerful Raycast extension providing native macOS integration:
- Unified Search: Fuzzy text, boolean expressions, and saved searches
- Quick Access: Launch via Raycast hotkey for instant prompt access
- Native Integration: Copy to clipboard, variable forms, metadata views
- Intelligent Detection: Automatically detects search type and routing
cd core/
go run main.go --tui # Test TUI interface
go run main.go --url-server # Test HTTP server
go test ./... # Run test suite
cd raycast-extension/
bun run dev # Development mode with live reload
bun run build # Build for production
bun run lint # Check code style
cd raycast-extension/
bun run build # Verify build succeeds
bun run publish # Submit to Raycast Store
The core application can be distributed as:
- Binary releases: Built executables for different platforms
- Go module:
go install
for Go developers - Package managers: Homebrew, apt, etc.
# Start server with smart git sync
cd core/ && ./pocket-prompt --url-server
# Health check
curl "http://localhost:8080/api/v1/health"
# List all prompts
curl "http://localhost:8080/api/v1/prompts"
# Search prompts
curl "http://localhost:8080/api/v1/search?q=ai"
# Boolean search
curl "http://localhost:8080/api/v1/boolean-search?expr=ai%20AND%20agent"
# Get specific prompt
curl "http://localhost:8080/api/v1/prompts/your-prompt-id"
# List saved searches
curl "http://localhost:8080/api/v1/saved-searches"
# Get prompts by tag
curl "http://localhost:8080/api/v1/tags/ai"
# API documentation
open "http://localhost:8080/api/docs"
Once installed, simply:
- Launch Raycast (⌘ + Space)
- Type "Search Prompts"
- Search your library with unified intelligent search
- Copy results instantly to clipboard
Contributions welcome for both components! See individual README files for component-specific guidelines:
MIT - See LICENSE files in individual component directories.