GPT Repository Loader is an open-source command-line utility that converts an entire code repository into a single, LLM prompt-friendly text format. Hosted on GitHub, the project has collected around 2,970 stars and was last actively updated in 2024. It was mostly built with GPT-4 itself, and it solves a very practical problem: how to hand a language model the full context of a codebase in one clean document.
The tool walks a repository’s file tree, respects ignore rules for files you do not want to include, and concatenates the contents into one structured text file with clear separators between files. That output can be pasted straight into ChatGPT, Claude, or any other LLM so the model can reason about the whole project at once. It is a small, focused script rather than a platform, which makes it easy to audit, fork, and adapt to your own workflow.
The repository includes a preamble template that explains the concatenated format to the model, plus an example workflow for summarizing and questioning large projects. Because everything lands in one text file, the technique works with web chat interfaces as well as APIs, and it remains a popular starting point that many later repo-to-prompt tools were inspired by. It is a good reminder that effective LLM tooling is often a small script with a well-designed output format.
GPT Repository Loader is completely free and open source under the MIT license. There is no hosted version or paid tier; you run it locally and only pay for whatever model API you feed the output into.
Developers who want an LLM to review, document, or debug an entire project at once, and prompt engineers who need a reliable way to package code as context.