Logswise CLI

A Rust-based command-line tool for note-taking with AI-powered suggestions and chat. Stores notes in Supabase and uses local Ollama models for AI features.

Current Features

• Note Storage: Store text notes (up to 10K characters) in Supabase with vector embeddings
• AI Suggestions: Get contextual suggestions based on stored notes using local LLM
• AI Chat: Interactive chat with AI that has access to your note history
• Interactive Mode: Menu-driven interface for all operations
• Personalization: Configure AI behavior and response styles
• Multiple Setup Modes: Express, template-based, and full configuration options

Technical Implementation

• Database: Supabase PostgreSQL with pgvector extension for semantic search
• AI Backend: Local Ollama models for text generation and embeddings
• Configuration: JSON-based config stored in ~/.logswise/setup.json
• Language: Rust with async/await for concurrent operations
• UI: Terminal-based with colored output and interactive menus

Installation

brew tap k61b/tap
brew install logswise-cli

Quick Setup

1. Install Ollama (if you haven't already):

   curl -fsSL https://ollama.ai/install.sh | sh
   ollama serve
   

2. Pull a model:

   # For AI features (chat/suggestions):
   ollama pull llama3
   
   # Or for search-only:
   ollama pull nomic-embed-text
   

3. Set up a Supabase project at supabase.com (free tier works)

4. Configure Logswise (Enhanced Setup Options):

   # Quick setup - only 5 questions, smart defaults
   logswise-cli setup --express
   
   # Use a template/preset for your role
   logswise-cli setup --template
   
   # Traditional full setup
   logswise-cli setup
   
   # Update existing configuration selectively
   logswise-cli setup --update
   

   Follow the prompts to enter your info, Supabase URL/key, and model name.

   💡 New: Express setup gets you running in under 2 minutes!

5. Initialize the database:

   logswise-cli init
   

6. Personalize your AI (optional):

   logswise-cli personalize setup
   

   This allows the AI to learn from your notes and improve suggestions.

Basic Usage

The CLI primarily operates in interactive mode. Simply run:

# Start interactive mode (default)
logswise-cli

# Or explicitly:
logswise-cli simple

This opens a menu-driven interface where you can:

• 💬 Chat with AI - Interactive AI chat session
• 📝 Add new log entry - Add notes to the system
• 💡 Get AI suggestions - Get context-aware suggestions
• 📊 View recent logs - Show recent note entries
• ⚙️ System & Configuration - Access system settings and diagnostics
• ❓ Help & Information - Access help documentation

Model Types

Embedding Models (search only):

• nomic-embed-text, bge-base-en, all-minilm
• Finds relevant notes but doesn't generate new text

LLMs (full AI features):

• llama3, deepseek-coder, mistral, phi3
• Generates suggestions and chat responses using your notes as context

Change models anytime by editing ~/.logswise/setup.json or re-running setup.

Available Commands

The CLI has a simple command structure focused on setup and interactive use:

# Setup and configuration
logswise-cli setup                    # Full interactive setup
logswise-cli setup --express          # Quick setup with defaults
logswise-cli setup --template         # Choose from role-based templates
logswise-cli setup --update           # Update existing configuration
logswise-cli setup --import <file>    # Import configuration from file

# Database initialization
logswise-cli init                     # Create/verify database schema

# Interactive mode (default)
logswise-cli                          # Main interactive interface
logswise-cli simple                   # Same as above (explicit)

# Personalization
logswise-cli personalize setup        # Configure AI personalization
logswise-cli personalize update       # Update personalization settings
logswise-cli personalize show         # View current personalization
logswise-cli personalize feedback     # Provide feedback on suggestions

# System info
logswise-cli --version

Note: Most functionality (adding notes, getting suggestions, chat, viewing recent entries, diagnostics, and stats) is accessed through the interactive mode menu system.

Troubleshooting

Common Issues:

• "embedding-only mode" message: You're using an embedding model. Switch to an LLM for chat/suggestions.
• Chat/suggestions don't work: Use the System Health Check in the interactive menu to diagnose. Check that Ollama is running and your model is available.
• Connection errors: Verify Ollama URL and Supabase credentials in ~/.logswise/setup.json

Health check: Access System Health Check through the interactive menu's System & Configuration section.

Contributing

See CONTRIBUTING.md for guidelines.

License

This project is licensed under the AGPL v3.0 License.