Title: Mini Claude - A Modular CLI Chatbot Inspired by OpenCode and Claude

Introduction: Mini Claude is a Python-based modular CLI chatbot that draws inspiration from OpenCode and Claude. This repository provides a basic structure and tools to get started with building your own conversational interfaces. Mini Claude features a separate memory layer, allowing for more advanced conversation management and context understanding.

Key Features:

• Modular architecture for easy customization and extension
• Separate memory layer for advanced conversation management and context understanding
• CLI-based interface for user interaction
• Inspired by OpenCode and Claude, with a focus on simplicity and ease of use

Getting Started: To set up the Mini Claude repository locally, follow these steps:

1. Clone the repository using git clone https://github.com/your-username/mini-claude.git
2. Install the required dependencies by running pip install -r requirements.txt
3. Navigate to the repository root directory and run python mini_agent.py to start the chatbot

Directory Structure: The repository is organized into the following directories:

• src/mini_claude/: contains the core chatbot logic and modules
• src/mini_claude/core/: contains the engine and provider modules for the chatbot
• src/mini_claude/tools/: contains utility functions for file management and terminal interactions
• src/mini_claude/memory/: contains the memory layer implementation for conversation management and context understanding

Configuration: To configure the chatbot, create a .mini_claude_config.json file in the repository root directory with the following format:

{
    "memory": {
        "size": 100,
        "expiration": 3600
    },
    "cli": {
        "prompt": "mini-claude> ",
        "welcome_message": "Welcome to Mini Claude! Type 'help' for commands."
    }
}

Replace the values with your desired settings for the memory layer and CLI interface.

Usage: To interact with the chatbot, simply run python mini_agent.py and follow the prompts. You can customize the chatbot's behavior by modifying the src/mini_claude/core/engine.py and src/mini_claude/core/provider.py files.

Memory Layer: The memory layer is a key feature of Mini Claude, allowing for advanced conversation management and context understanding.

1. The Short-Term Memory (session.py) This handles the immediate conversation history.

Persistence: Every message (User, Assistant, or Tool result) is saved to .opencode/sessions/latest.json the moment it happens.

Auto-Resumption: When the engine starts, it calls session.load(), allowing you to pick up exactly where you left off.

Context Cleaning: We implemented a "Strip" logic that removes extra SDK fields (like annotations) before saving, ensuring the Groq API doesn't reject the history later.

2. The Context Manager (compact_context) This is the "Memory Optimization" logic located within the engine.

Sliding Window: It monitors the history length. Once it hits your max_history (15 turns), it triggers a compaction.

Summarization: It uses a smaller, faster model (Llama-3.1-8b) to turn the middle of the conversation into a single "Technical Status Report," saving thousands of tokens while keeping the "Goal" and "Recent Actions" intact.

3. The Long-Term Memory & Grounding (storage.py) This is the "Project Knowledge" layer.

Root Detection: It identifies the project root so the AI knows its boundaries.

CLAUDE.md Integration: This acts as the "Permanent Memory." It’s where the AI reads its custom project instructions and coding standards every time you run a command.

Environment Mapping: It provides the "Repo Map" to the system prompt, so the AI always knows the current file structure without you having to explain it.

Contributing: We welcome contributions to the Mini Claude repository! If you'd like to add a new feature or fix a bug, please submit a pull request with a clear description of your changes.

License: Mini Claude is licensed under the MIT License. See LICENSE for details.