Gitingest turns any GitHub repository into a single, prompt-friendly text digest that you can paste straight into an LLM. The open-source tool has more than 15,900 stars on GitHub and was actively updated in 2026.
The trick is memorable: take a GitHub URL and replace “hub” with “ingest” to get gitingest.com, which returns the repository’s directory tree, file contents and summary statistics formatted for a model’s context window. It is the fastest way to ask ChatGPT, Claude or another assistant questions about an unfamiliar codebase, prepare code for review, or feed a project into an agent. Beyond the hosted site, Gitingest also runs as a Python package and command-line tool for local or private use.
A few thoughtful details explain its popularity. The digest leads with summary statistics โ file counts and an estimated token count โ so you know whether the result will fit your model’s context window before pasting it. You can scope the output to a specific branch or subdirectory, cap individual file sizes, and filter with include/exclude patterns, which keeps huge monorepos manageable. Because the same engine runs behind the website, the Python package and the CLI, teams can prototype in the browser and then automate the identical extraction inside their own agent or documentation pipelines.
Gitingest is free and open source under the MIT license, and the hosted gitingest.com service is free to use for public repositories.
Developers who paste code into AI chats, reviewers exploring unfamiliar projects, and agent builders who need a simple codebase-ingestion step.