LLM Code Context

🚀 LLM Context

LLM Context is a tool that enables developers to quickly inject project content into the chat interface of large language models (LLMs). It simplifies the development process by intelligently selecting files using the .gitignore pattern and offers both command - line and direct integration into LLMs.

🚀 Quick Start

Integration with Claude Desktop

Add the following configuration to package.json:

"llm-context": {
  "rules": [
    "src/**/*",
    "!src/**/*.ts"
  ],
  "outlines": [
    "src/**/*",
    "!src/**/*.ts"
  ]
}

Then run the initialization command:

llm-context init

✨ Features

• Intelligent File Selection: Filters project files using the .gitignore pattern.
• Multi - rule Configuration: Supports system rules and custom rules, suitable for different development scenarios.
• Code Navigation Function:
  1. Intelligent Code Overview: Generates an overview of the code structure and highlights key definitions.
  2. Definition and Implementation Extraction: Quickly obtain the code implementation for LLM requests through the llm-context clip-implementations command.
• Advanced Customization:
  • Offers flexible template and prompt functions to meet personalized needs.
  • Supports file change detection for easy tracking of project updates.

📦 Installation

To install LLM Context, follow these steps:

1. Initialize project configuration:

   llm-context init
   

2. Set rules (optional):

   llm-context set-rule <n>
   

   System rules start with the lc- prefix, and users can create custom rules as needed.

3. Select files:

   llm-context sel-files
   

4. Generate context:

   llm-context context [-p] [-u] [-f FILE]
   

   • -p: Include prompt instructions.
   • -u: Include user comments.
   • -f FILE: Output to the specified file.

💻 Usage Examples

Basic Usage

The basic steps for using LLM Context are as follows:

# Initialize project configuration
llm-context init
# Select files
llm-context sel-files
# Generate context
llm-context context

Advanced Usage

LLM Context provides rich customization options. For example, you can use the following command to generate context with prompt instructions and output it to a specific file:

llm-context context -p -f output.txt

📚 Documentation

For a comprehensive list of similar tools, please refer to comprehensive list of alternatives .

🔧 Acknowledgments

LLM Context inherits the innovative spirit of previous generations of AI - assisted development tools:

• It is inherited from LLM Code Highlighter , a TypeScript library for IDE integration.
• Its concept originated from RubberDuck and subsequent contributions to Continue .
• It benefits from the assistance of GPT - 4 and Claude - 3.5 - Sonnet in project development.

📄 License

This project is licensed under the Apache License 2.0. Please refer to the LICENSE file for specific terms.