Ex-GitHub CEO Dohmke Launches Entire, an Agent-First Git Forge

Former GitHub CEO Thomas Dohmke launched Entire, a Git forge with agent-run code review, faster hosting, and agent memory, built for AI coding agents.

@sharedot

Ex-GitHub CEO Dohmke Launches Entire, an Agent-First Git Forge

By @sharedot · · 8 pages

  • Programming
  • Developer Tools
  • AI Coding Agents
  • Git

Former GitHub CEO Thomas Dohmke launched Entire, a Git forge with agent-run code review, faster hosting, and agent memory, built for AI coding agents.

What happened: Entire goes live for humans and AI agents

On October 26, Thomas Dohmke, who left GitHub as its CEO over a year ago, unveiled Entire, what the company calls the first Git forge built for humans and AI agents to work together.

The platform lets developers host, manage, and collaborate on code, and it is available to try today for all developers. It is positioned as a GitHub alternative that helps teams merge changes as fast as AI coding agents generate them.

The launch builds on Checkpoints, Entire's February product that captures agent prompts, decisions, and session history alongside code changes, and lands just days before GitHub's annual developer conference on October 28-29.

Why it's surprising: the pull request itself gets replaced

The surprising move is that Entire does not just clone GitHub's workflow; it replaces the pull request, the model that transformed software development, with a new construct called Trails.

Entire presents Trails as a heads-up display similar to an airline pilot's, with customized evals scored by agents on every code push. Dohmke said the idea, the agent trace, and the outcome of a developer's sessions are treated as first-class artifacts for review alongside the code.

The timing is pointed: it arrives as GitHub has admitted agent pull requests are saturating review bandwidth, and days before its own developer conference.

The evidence: three flagship features and benchmark claims

Entire launched with three features. First, Trails lets AI agents automatically check code changes on every push, address findings in a loop, and let developers merge once all evals pass, viewable via the web interface or the open-source CLI.

Second, distributed Git hosting that the startup claims pushes code up to 89 times faster than GitHub and six times faster than Cursor Origin, with private repositories hostable in India, the United States, Europe, and Australia.

Third, Continuous Memory, which stores past agent sessions in the repo; Entire claims it can reduce token usage by up to 71 percent.

The stakes: bottlenecks shift from writing to reviewing code

Depot co-founder and CEO Kyle Galbraith said the biggest bottleneck in software development has shifted from writing code to the processes that follow, such as building and testing it, and that Depot CI helps developers and AI agents speed those up while keeping existing GitHub Actions workflows.

Entire is partnering with Buildkite and Depot so builds trigger on new pushes and results appear in Trails.

Dohmke also sought to ease trust concerns, telling ET Enterprise AI that Entire does not use customer source code to train any AI coding model, and that its open-source CLI redacts secrets and personal data locally before anything is stored.

The backers: $60 million seed at a $300 million valuation

Entire raised $60 million in seed funding in February at a $300 million valuation, in a round led by Felicis with participation from Madrona, M12, and Basis Set, according to Moneycontrol.

Its backers also include Y Combinator CEO Garry Tan, Datadog CEO Olivier Pomel, and former Yahoo CEO Jerry Yang. The startup has grown to a global team of 50, including employees who have worked at GitHub, Microsoft, Atlassian, and DigitalOcean.

The launch sits on the decentralized Git network Entire announced in July, which ZDNET notes was designed to withstand the surge in AI coding activity behind mounting GitHub outages.

What comes next: migration friction versus agent-era tooling

Adoption remains the open question.

ET Enterprise AI notes GitHub dominates with a vast developer community and established workflows, and that moving to Entire means migrating repositories, preserving code history, and rebuilding integrations, which is especially challenging for enterprises with entrenched collaboration processes.

Entire is currently focused on teams using AI agents for software development and plans to expand to open-source maintainers and large enterprises.

On the roadmap, Moneycontrol reports future features including public repositories, self-hosted Git nodes, live-streamed sessions, and more integrations with other developer tools.

Sources

  1. moneycontrol.com › Former GitHub CEO Thomas Dohmke launches rival platform built for humans and AI coding agents
  2. zdnet.com › Github competitor Entire is live now – how the agent-first platform works
  3. enterpriseai.economictimes.indiatimes.com › Former GitHub CEO's Entire launches Git Forge for AI agents

More on Programming