ContextR is a tool that helps developers collect and package code files for use with Large Language Models (LLMs) like GPT-4. It makes it easier to provide LLMs with the right context about your codebase, leading to more accurate and helpful AI responses.
We've added a plugin system that makes ContextR much more extensible. This means:
- For Users: More features and customization options without waiting for core updates
- For Developers: Ability to create custom extensions for specific needs
- For the Business: A more adaptable product that can grow with changing requirements
We've added features to help protect sensitive information:
- Automatic Detection: ContextR can now identify potential API keys, passwords, and other sensitive data
- GitIgnore Integration: Files that are excluded from version control are automatically excluded from context
- Env File Handling: Special handling for environment files to include only keys without their values
New visualization options make it easier to understand and work with code context:
- Tree View: See the full structure of your project
- Studio UI: A visual interface for building context and managing files
- HTML & Markdown Output: Generate well-formatted documentation from your code context
Direct integration with language models for code analysis:
- Local LLM Support: Works with locally installed models (no API keys required)
- Code Review: Automated code review and summarization
- Security Analysis: Identify potential security issues in your code
Enhanced CLI with more features and better usability:
- New Commands: More options for working with your code
- Better Help: Improved documentation and examples
- Simplified Workflow: Common tasks are now easier to perform
- Improved Developer Productivity: Developers can more easily provide context to LLMs, leading to better AI assistance
- Enhanced Security: Reduced risk of accidentally exposing sensitive information to LLMs
- Better Collaboration: Visual tools and improved output formats make it easier to share context with team members
- Future-Proofing: The plugin system allows for adaptation to new LLM technologies and requirements
- No Breaking Changes: Fully backward compatible, so existing users can upgrade without disruption
- VSCode Extension: We're planning to develop a VSCode extension for even tighter integration with the development workflow
- More Plugins: Additional security scanners, output renderers, and LLM integrations
- Enhanced Documentation: More examples and guides for using the new features
We welcome feedback on the new features and suggestions for future improvements. The development team is available to answer questions and provide support for the new version.