Former GitHub Chief Thomas Dohmke Takes On His Old Company With Entire
Trending Topics Jakob Steinschaden
Thomas Dohmke just launched Entire, a GitHub rival built for AI agents. It bets code review will shift from humans reading every line to machines checking the work.
Based on reporting by Trending Topics, Jakob Steinschaden — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Thomas Dohmke has gone from GitHub’s last chief executive to one of its newest challengers. His new company, Entire, is pitched as a Git forge for a world where AI agents do most of the coding and humans supervise the output.
The company says its main idea is to replace the old pull request with something called a Trail. Instead of asking people to inspect every line, Entire has several AI agents score each push against custom checks, including security and “slop,” its term for low-quality AI-generated code. The results are shown together for developers, and once the criteria are met, the code can be merged without a full human line-by-line review.
Entire is also building its own Git-compatible hosting. The company claims pushes are up to 89 times faster than on GitHub and six times faster than on Cursor Origin, and it lets private repositories live in the United States, the European Union, India or Australia. That regional choice is meant to matter for data protection, especially in Europe.
There’s a memory layer too. Coding agent sessions are stored in the repository, searchable by other agents, while a codebase graph is supposed to cut token use by up to 71 percent. Entire says an “agentic brain” will also help keep the same mistakes from being repeated. It is partnering with Buildkite and Depot so teams can bring over build and test pipelines from GitHub Actions.
Dohmke framed the shift bluntly: the pull request changed open source and then enterprise software, but now agents are writing almost all of the code, and they’re doing it faster and in more complex ways. Entire raised $60 million in February at a $300 million valuation, with backing from Felicis, Madrona, M12, Cherry Ventures, Picus Capital and Global Founders Capital, plus angel investors including Jerry Yang and Olivier Pomel. The team has since grown to 50 people. GitHub, meanwhile, has been dealing with hundreds of incidents and more than 17 million AI-generated pull requests a month. Entire is starting with AI-native teams first, while public repositories, self-hosting and broader enterprise and open-source support are still ahead.
My take — AI-written commentary, not fact-checked reporting
This is the obvious next fight: if AI writes the code, the whole ritual around code review has to change. GitHub built the world around human attention, and that assumption is already wobbling under the weight of agent spam and platform outages. The funny part is that the replacement pitch is still old-school software discipline, just with more machines and less ceremony.
Read more about this at: Trending Topics