Ex-GitHub CEO Dohmke launches Entire to host AI-generated code

Key Takeaways

• Former GitHub CEO Thomas Dohmke launched Entire, a Git hosting platform designed for AI coding agents
• Entire claims 2.1 million pushes per hour compared to Cursor Origin's 81,000
• The platform mirrors GitHub repos and tracks AI agent decision-making for auditing

Thomas Dohmke, who led GitHub until recently, has launched Entire, a Git hosting network built from scratch for AI coding agents. The startup addresses a real problem: GitHub's infrastructure is buckling under the load of AI-generated commits, and traditional hosting platforms were never designed for machines that code.

Entire.io, now open to waitlisted developers in the US, EU, and Australia, lets users mirror their public or private GitHub repositories. The pitch is simple: let AI agents churn through code on Entire's infrastructure instead of straining GitHub resources needed for production deployments.

Why GitHub can't keep up with vibe coding

"Vibe coding" refers to developers describing what they want in natural language and letting AI handle implementation. This sounds efficient until you realize what it does to Git infrastructure. AI agents make far more commits, branches, and API calls than humans. They don't take coffee breaks. They don't sleep. And when thousands of them hammer a centralized platform simultaneously, that platform struggles.

Dohmke knows this firsthand. He ran GitHub during the explosive growth of GitHub Copilot and watched AI tooling reshape developer workflows. Now he's building what GitHub probably wishes it had: infrastructure that treats AI agents as first-class citizens rather than unusually prolific humans.

“The question is not if Git survives through the sheer weight of its ecosystem lock-in. The question is how we can expand, rewire, and evolve Git hosting for a world where AI agents are the primary producers of code.”

— Thomas Dohmke, Founder of Entire

Entire's performance claims vs. Cursor Origin

SpaceX's Cursor launched its own GitHub alternative, Cursor Origin, last month. Entire is positioning itself directly against it with aggressive benchmarks.

Those numbers suggest Entire can handle roughly 25 times more pushes per hour than Cursor Origin. The company attributes this to its architecture: decentralized by design, optimized for low latency, and built to manage regional content distribution. Whether these benchmarks hold under real-world AI agent traffic remains to be seen.

The Entire CLI: tracking why AI made changes

Beyond hosting, Entire ships a CLI tool that integrates with Git to capture AI agent sessions alongside commits. It records prompts, responses, file changes, and context. The goal is auditability: not just seeing what an AI changed, but understanding the reasoning behind those changes.

For enterprises, this matters. When an AI agent introduces a bug or a security vulnerability, someone has to trace back through its decision chain. Traditional Git history shows diffs. Entire's approach shows intent.

This also hints at Entire's business model. Git hosting can be commoditized. Agent auditing, compliance reporting, and insights into AI decision-making patterns are services companies will pay for. The "data exhaust" from AI coding sessions has value, and Entire is positioning itself to capture it.

Decentralization was Git's design all along

Dohmke acknowledges an irony here. Git, the 21-year-old version control system created by Linus Torvalds, was always meant to be decentralized. Every clone contains a complete repository and its full history. The protocol doesn't require a central server.

In practice, GitHub, GitLab, and Bitbucket funneled developers into centralized platforms. That worked fine when humans were the primary committers. It breaks down when AI agents generate code at machine speed.

Entire plans to open-source its Git network and allow self-hosting in the coming months. "We will continue building up and down the stack towards an open, decentralized, independent developer ecosystem for any agent and any human," Dohmke said.

What this means for enterprise IT

If your engineering teams use AI coding assistants heavily, you're probably already seeing the effects: slower clone times, rate limiting, and occasional outages on GitHub. Entire offers a pressure valve. Mirror your repos, let agents work on Entire's infrastructure, and merge back when ready.

The agent auditing features are also worth watching. As AI-generated code becomes a larger portion of codebases, compliance and security teams will demand better visibility into how that code was produced. Entire's CLI could become a required tool for regulated industries.

Frequently Asked Questions

What is Entire and who founded it?

Entire is a Git hosting network designed for AI coding agents, founded by Thomas Dohmke, the former CEO of GitHub. It launched in July 2026 and is currently available to waitlisted developers in the US, EU, and Australia.

How does Entire differ from GitHub?

Entire is built specifically for AI agent traffic patterns, claiming 2.1 million pushes per hour compared to traditional platforms. It also includes tooling to track AI decision-making for auditing purposes, not just code changes.

Can I migrate my GitHub repositories to Entire?

Entire currently supports mirroring public or private GitHub repos. You can also use Entire-native branches exclusively. Full migration tools have not been detailed yet.

Will Entire be open source?

Yes. The company plans to open-source its Git network and enable self-hosting in the coming months.

What is vibe coding?

Vibe coding refers to using AI agents to write code based on natural language descriptions, with minimal human oversight. It generates significantly more commits and API calls than traditional development.

Source: www.theregister.com