This framework addresses the complexity of managing context, dependencies, and tasks in large-scale projects, specifically within the VS Code environment using the Cline extension. It targets developers and researchers working on intricate projects, offering a structured and efficient way to maintain project state and improve workflow.

How It Works

CRCT employs a recursive, file-based approach with a modular dependency tracking system. It utilizes "Contextual Keys" (KeyInfo) for precise, hierarchical dependency tracking and aggregates these dependencies for a comprehensive project view. The system stores context, instructions, and outputs persistently using the VS Code file system, ensuring state integrity through a Mandatory Update Protocol. It supports configurable embedding devices (CPU, CUDA, MPS) and file exclusion patterns for optimized analysis.

Quick Start & Requirements

• Install dependencies: pip install -r requirements.txt
• Setup: Clone the repo, install requirements, copy cline_docs/prompts/core_prompt(put this in Custom Instructions).md into the Cline system prompt field in VS Code.
• Start: Type Start. in the Cline input.
• Prerequisites: VS Code with the Cline extension installed. Manual installation of CUDA or MPS drivers may be required for GPU acceleration.
• Links: Project Repository

Highlighted Details

• Achieves significant time savings with caching and batch processing (e.g., 30 seconds vs. 11 minutes in tests).
• Introduces show-dependencies for simplified dependency inspection and visualize-dependencies (experimental) for generating Mermaid dependency flowcharts.
• Allows manual linking of external dependencies to local files, useful for integrating documentation with code.
• Supports file exclusion patterns for customized project analysis.

Maintenance & Community

The project acknowledges contributions from biaomingzhong for prompt integration. Feedback and bug reports are encouraged via GitHub Issues.

Licensing & Compatibility

The project is licensed under the MIT License. This license is permissive and generally compatible with commercial use and closed-source linking.

Limitations & Caveats

The visualize-dependencies command is experimental and subject to change. The system does not automatically install CUDA or MPS requirements, necessitating manual installation for GPU acceleration.