ck:repomix

Pack a codebase into a single LLM context file by consolidating repository contents into XML, Markdown, JSON, or plain text snapshots that preserve file hierarchy and metadata. This process formats local folders or remote Git repositories for AI consumption.

Yes, you can package remote GitHub repositories via npx, optionally targeting specific commits. This allows you to prepare third-party or large codebases for LLM security audits and codebase comparisons without cloning them locally first.

Count tokens per file and produce token trees to manage LLM context limits effectively. You can also include or exclude specific files using globs and .repomixignore rules to ensure the packaged snapshot fits your model's context window.

Yes, creating code snapshots with repomix respects .gitignore rules, along with custom .repomixignore configurations. This ensures sensitive files are properly ignored and excluded from the final AI-ready output before sharing it with LLMs.

Automate repository packaging for multiple projects using batch processing scripts that enable environment-based configuration, progress reporting, and CI integration. This allows you to systematically generate AI-friendly snapshots across various repositories.

Yes, you can remove comments and perform secret scanning before sharing a codebase snapshot. These security-aware processing options help sanitize the AI-ready output, ensuring sensitive information is stripped from the consolidated file.