LLLM - Life Long Learning Model

An AI-powered CLI learning assistant that helps you learn, retain, and test your knowledge using Claude.

Features

• Ask Mode: Interactive Q&A with Claude, automatically organized by topic
• Quiz Mode: Test your knowledge with AI-generated open-ended questions
• Smart Topic Detection: Automatically categorizes your questions or asks for clarification
• Adaptive Feedback: Get hints and detailed critiques on your quiz answers
• Progress Tracking: (Coming soon) Track your learning progress over time

Installation

1. Clone the repository
2. Install dependencies:

bun install

3. Link the CLI globally:

bun link

4. Set your Anthropic API key:

cp .env.example .env
# Edit .env and add your API key

Usage

Ask Mode

Ask questions and learn interactively. Your conversations are automatically saved and organized by topic.

lllm ask

Example session:

> What is a channel in Go?
[Topic: golang]

[Claude's response with streaming...]

> What does panic mean in Go?
[Topic: golang]

[Claude's response...]

Quiz Mode

Test your knowledge on topics you've learned about.

lllm quiz

Features:

• Choose specific topics or quiz on all topics
• Open-ended questions with difficulty ratings
• Up to 3 attempts per question with hints
• Detailed feedback on your answers

Configuration

View your current configuration:

lllm config

Your API key is stored in a .env file in the project directory:

1. Copy the example: cp .env.example .env
2. Edit .env and add your Anthropic API key
3. Get your API key from: https://console.anthropic.com/settings/keys

The .env file is gitignored and will not be committed to version control.

How It Works

1. Learning: When you ask questions in ask mode, Claude responds and saves the Q&A to topic-specific markdown files in src/.storage/.topics/

2. Topic Detection: The system intelligently determines which topic your question belongs to. If ambiguous, it will ask for clarification.

3. Quizzing: Quiz mode reads your saved learning materials and generates contextual questions to test your understanding.

4. Adaptive Feedback: When answering quiz questions, you get:

   • Immediate feedback on correctness
   • Hints if your answer is incomplete (up to 3 attempts)
   • Detailed explanations of what you missed

Project Structure

lllm/
├── .env.example          # Example environment variables
├── .gitignore           # Git ignore rules
├── bun.lock            # Bun lock file
├── package.json        # Project dependencies and scripts
├── tsconfig.json       # TypeScript configuration
├── README.md           # Project documentation
├── LICENSE             # MIT license
└── src/
    ├── commands/
    │   ├── ask.ts      # Ask mode implementation
    │   ├── quiz.ts     # Quiz mode implementation
    │   └── config.ts   # Configuration command
    ├── lib/
    │   ├── claude.ts         # Claude API wrapper
    │   ├── topics.ts         # Topic management & storage
    │   └── quiz-generator.ts # Quiz generation logic
    ├── .storage/             # (gitignored - created on first run)
    │   ├── .topics/          # Markdown files per topic
    │   ├── .metadata.json    # Topic metadata
    │   └── progress.json     # Quiz progress tracking
    └── index.ts              # CLI entry point

Future Enhancements

• Full progress tracking implementation
• Spaced repetition reminders
• Export to Anki flashcards
• Web interface
• Multiple AI model support
• Collaborative learning mode

Development

Run without linking:

bun run src/index.ts <command>

Build for distribution:

bun build

Tech Stack

• Runtime: Bun
• CLI Framework: Commander.js
• AI: Anthropic Claude API (Sonnet 4.5)
• Language: TypeScript

License

MIT

---

Built with Claude Code 🤖