📌Overview

Purpose:
LLM Context is designed to help developers inject relevant project content into Large Language Model chat interfaces efficiently and intelligently.

Overview:
LLM Context streamlines integrating code and text project information into LLM-based chat systems. It supports both direct integration with advanced chat platforms (via Model Context Protocol) and works universally through a command-line clipboard workflow. By leveraging .gitignore for smart file selection and customizable rule-based profiles, it ensures only the most pertinent project data is included, enhancing AI-assisted development and collaboration.

Key Features:

• Smart File Selection: Automatically selects relevant files for LLM context based on .gitignore patterns, minimizing noise and focusing on essential project content.

• Seamless Integration & Workflows: Offers both native integration with platforms like Claude Desktop (via MCP) and universal CLI/clipboard workflows, making it compatible with a wide range of LLM chat interfaces and developer environments.

---

LLM Context

LLM Context is a tool to help developers inject relevant content from code or text projects into Large Language Model (LLM) chat interfaces. It uses .gitignore patterns for smart file selection and provides both a streamlined command-line workflow and direct LLM integration through the Model Context Protocol (MCP).

What's New in v0.3.0

Configuration now uses a Markdown (with YAML front matter) rules system, replacing the previous TOML/YAML profiles. This is a breaking change. See the User Guide for details.

Why Use LLM Context?

For more background on LLM Context’s approach to AI-assisted development, see:

• LLM Context: Harnessing Vanilla AI Chats for Development
• Full Context Magic - When AI Finally Understands Your Entire Project

Usage Patterns

• Direct LLM Integration: Native support for Claude Desktop via the MCP protocol.
• Chat Interface Support: Works with any LLM chat interface via CLI/clipboard; optimized for persistent context interfaces like Claude Projects and Custom GPTs.
• Supported Projects: Suitable for code repositories and collections of text, markdown, or HTML documents.
• Project Size: Best for projects that fit within the LLM context window (large project support is upcoming).

Installation

With uv:

uv tool install "llm-context>=0.3.0"

To upgrade:

uv tool upgrade llm-context

Warning: LLM Context is under active development. Updates may overwrite configuration files prefixed with lc-. Version control your configuration files.

Quickstart

Using MCP with Claude Desktop

Add this to your claude_desktop_config.json:

{
  "mcpServers": {
    "CyberChitta": {
      "command": "uvx",
      "args": ["--from", "llm-context", "lc-mcp"]
    }
  }
}

Then, either:

1. Say: "I would like to work with my project" (Claude will prompt for the project root), or
2. Specify: "I would like to work with my project /path/to/your/project"

For more, see Full Context Magic.

CLI Workflow

1. Navigate to the project root.
2. Initialize: lc-init (once).
3. Select files: lc-sel-files.
4. (Optional) Review your selection in .llm-context/curr_ctx.yaml.
5. Generate context: lc-context (add -p for prompts, -u for user notes).
6. Paste context into your LLM’s knowledge section or chat interface.
7. If the LLM requests additional files:
   • Copy the file list from the LLM.
   • Run lc-clip-files.
   • Paste contents back to the LLM.

Core Commands

• lc-init: Initialize project configuration
• lc-set-rule <n>: Switch rules
• lc-sel-files: Select files for context
• lc-sel-outlines: Select files for outline generation
• lc-context [-p] [-u] [-f FILE]: Generate and copy context
  • -p: Include prompt instructions
  • -u: Include user notes
  • -f FILE: Output to file
• lc-prompt: Generate project instructions for LLMs
• lc-clip-files: Process LLM file requests
• lc-changed: List recently modified files
• lc-outlines: Generate code outlines
• lc-clip-implementations: Extract code requested by LLMs (except C/C++)

Features & Advanced Usage

• Smart file selection via .gitignore patterns
• Rule-based context profiles (system and user-defined)
• Code Navigation:
  1. Smart Code Outlines: Automatic, high-level codebase structure summaries
  2. Implementation Extraction: Returns specific implementations the LLM asks for using lc-clip-implementations
• Customizable templates and prompts

See the User Guide for documentation.

Similar Tools

Explore alternative tools in this comprehensive list.

Acknowledgments

LLM Context builds on prior AI-assisted development tools:

• Successor to LLM Code Highlighter.
• Inspired by work on RubberDuck, Continue, and Aider Chat.
• Uses tree-sitter tag query files from Aider Chat.

Thanks to the open-source community and LLMs (including Claude-3.5-Sonnet) for their support in this project's development.

License

Licensed under the Apache License, Version 2.0. See the LICENSE file for details.