Skip to main content
Former GitHub CEO, Nat Friedman, presents "Entire," a new Git platform designed for managing AI agents.

Editorial illustration for Former GitHub CEO Launches Entire, a Git Platform for AI Agents

GitHub's Ex-CEO Launches Entire, New Git Platform for AI

Former GitHub CEO Launches Entire, a Git Platform for AI Agents

• 4 min read

Thomas Dohmke ran GitHub until he stepped down earlier this year, and he's back with a new company called Entire. On October 26, the startup launched Git Forge, a developer platform built on the premise that Git itself needs rethinking now that AI agents, not just humans, are writing and reviewing code.

The platform ships with three pieces: Trails, an agent-based review system; distributed Git hosting with servers in the US, Europe, Australia and India; and Continuous Memory, which saves past coding sessions and prompts inside repositories so agents can pull up that context later instead of starting from scratch. Entire says its source control server pushes code up to 89 times faster than GitHub and six times faster than Cursor Origin, numbers that come from the company's own benchmarks rather than outside testing.

The bigger pitch is about what happens to code review once agents start generating more output than any developer could reasonably read line by line. Pull requests were built for a world where a human wrote the diff and another human judged it on comments and CI results. Entire argues that model breaks down fast when the author and the reviewer are both software.

Entire, a startup founded by former GitHub CEO Thomas Dohmke, has launched a Git-based developer platform that uses AI agents to review code, manage repositories and retain context from previous coding sessions.

Why this matters

Dohmke built his career on the assumption that Git stays human-centric and GitHub just wraps around it. Entire bets the opposite: that agents writing and reviewing code at scale need their own substrate, not a layer bolted onto infrastructure designed for human commits. Continuous Memory is the piece worth watching closest.

If coding sessions and prompts persist in the repository itself, that changes what "context" means for an AI agent working across weeks or handoffs between tools like Claude Code and Codex. For founders building agent-heavy workflows, the pitch to host directly or mirror from GitHub lowers switching costs, which matters more than any single feature. For developers, the real test is whether Trails produces review quality that survives contact with messy, real codebases, not demo repos.

Leaning on Daytona and E2B for isolated cloud execution also means Entire's reliability depends on vendors most teams have never evaluated for production load. Worth tracking: whether GitHub itself ships comparable memory and review tooling before Entire proves out at scale.

Common Questions Answered

What is Git Forge and what are its main components?

Git Forge is a developer platform launched by Entire that is specifically designed for AI agents to write and review code. The platform includes three key components: Trails, an agent-based review system; distributed Git hosting with servers across the US, Europe, Australia, and India; and Continuous Memory, which preserves context from previous coding sessions.

How does Entire's approach differ from GitHub's traditional model?

While GitHub was built around the assumption that Git remains human-centric with GitHub as a wrapper around it, Entire operates on the opposite premise: that AI agents writing and reviewing code at scale need their own dedicated substrate rather than a layer added to infrastructure designed for human commits. This fundamental difference reflects a shift in how version control systems need to be architected for AI-driven development.

What is Continuous Memory and why is it significant for AI agents?

Continuous Memory is a feature that saves and persists coding sessions and prompts directly within the repository itself, changing what context means for AI agents working across extended periods or during handoffs between different agents. This capability allows AI agents to maintain continuity and understanding across weeks of development work, rather than starting fresh with each session.

Who founded Entire and what is his background?

Entire was founded by Thomas Dohmke, who previously served as the CEO of GitHub until he stepped down earlier this year. His extensive experience leading GitHub gives him deep expertise in version control systems and developer platforms, which informs Entire's approach to building infrastructure for AI agents.

LIVE09:10Sarvam Signs 5-Year Defence AI Deal Covering Army, Navy and Air Force