关于
Codefetch simplifies the process of preparing code for analysis by Large Language Models (LLMs). It fetches code files from a specified directory, respecting .gitignore and .codefetchignore, and outputs them into a single Markdown file, complete with line numbers. Features include filtering by file extension, including or excluding specific files/directories, generating a project tree, and adding prompts for specific instructions to the LLM. Token limiting strategies and various token counting methods are also supported to optimize output for different AI models.
主要功能
- Supports filtering files by extension and excluding directories.
- Includes prompt support for guiding LLM analysis.
- Offers token limiting strategies for large codebases.
- 266 GitHub stars
- Converts code to Markdown format with line numbers.
- Generates a visual project tree representation.
使用案例
- Preparing code for analysis and review by LLMs.
- Generating Markdown documentation from a codebase.
- Integrating code snippets into AI-powered tools like bolt.new and cursor.com.