How does Repomix handle large repositories that exceed token limits?

It uses segmented bundling and a --compress flag to reduce token usage, allowing you to load only the specific modules or features relevant to your current task.

What bundle formats does Repomix support?

It supports XML, which is optimized for LLM processing, Markdown for human readability, and JSON for programmatic use cases.

Is my code secure when using the Repomix skill?

Yes, Repomix operates entirely on your local files and includes a security check powered by secretlint to ensure sensitive information isn't included in the bundles.

How do I refresh my project context after making changes?

You can simply run the /repomix bundle or /repomix pack command to regenerate your bundles and provide Claude with the most up-to-date version of your code.

What is Repomix used for in Claude Code?

Repomix packs your repository into a single, structured file so Claude can understand your entire project's context, architecture, and logic at once.

Repomix Codebase Bundler

Name: Repomix Codebase Bundler
Author: teslasoft-de

byteslasoft-de

생산성 및 워크플로

Packs complex codebases into AI-optimized XML and Markdown bundles to provide Claude with comprehensive project context.

소개

Repomix is a specialized utility for Claude Code designed to bridge the gap between large repositories and AI context windows by condensing codebases into structured, LLM-friendly formats. It features intelligent auto-discovery of project types, token-saving compression, and segmented bundling, which allows developers to selectively load specific modules or the entire project architecture. This skill is essential for performing deep-dive debugging, comprehensive code reviews, and architectural analysis where cross-file context is critical for accurate AI reasoning.

주요 기능

Multi-format support including XML (best for LLMs), Markdown, and JSON
0 GitHub stars
Intelligent auto-discovery of project structures and file types
Built-in security scanning via secretlint to prevent leaking sensitive data
Token-optimized compression and segmented bundling to stay within context limits
Simplified CLI commands for status checks and rapid bundle regeneration

사용 사례

Mapping and understanding the architecture of unfamiliar or legacy projects
Creating portable codebase snapshots for collaboration and context-sharing in new sessions
Providing comprehensive context for AI-driven code reviews and debugging

소개

주요 기능

Multi-format support including XML (best for LLMs), Markdown, and JSON
0 GitHub stars
Intelligent auto-discovery of project structures and file types
Built-in security scanning via secretlint to prevent leaking sensitive data
Token-optimized compression and segmented bundling to stay within context limits
Simplified CLI commands for status checks and rapid bundle regeneration

사용 사례

Mapping and understanding the architecture of unfamiliar or legacy projects
Creating portable codebase snapshots for collaboration and context-sharing in new sessions
Providing comprehensive context for AI-driven code reviews and debugging