coderamp-labs/gitingest

What it solves

Providing LLMs with the full context of a codebase is often difficult because repositories are spread across many files and directories. Gitingest simplifies this by consolidating a Git repository or local directory into a single, prompt-friendly text digest that can be easily pasted into an AI model.

How it works

The tool analyzes a Git repository (via URL) or a local directory and extracts the relevant code. It then formats this content into a structured text dump, including a directory tree and statistics such as token count and total size. It can be used as a web service (gitingest.com), a CLI tool, a Python package, or a browser extension.

Who it’s for

Developers who want to feed entire codebases or specific subdirectories into LLMs for analysis, debugging, or feature implementation without manually copying and pasting multiple files.

Highlights

• Multiple interfaces: Available as a website, CLI, Python library, and browser extensions for Chrome, Firefox, and Edge.
• Smart formatting: Specifically optimized for LLM prompt consumption.
• Contextual statistics: Provides file/directory structure and token counts using tiktoken.
• Flexible ingestion: Supports private repositories via GitHub Personal Access Tokens and can include submodules or git-ignored files.
• Self-hostable: Can be deployed via Docker or Docker Compose with S3-compatible storage support.