Former GitHub CEO launches distributed Git network for AI agents

Entire, a new platform from former GitHub CEO Thomas Dohmke, has launched a preview of a distributed Git network designed to handle the traffic of AI coding agents. The network lets developers mirror existing GitHub repositories to regional nodes, offloading heavy read traffic from central servers.

What it does

The preview allows one-step mirroring of GitHub repos, with active regions in the US, EU, and Australia. Agents clone and pull from the regional Entire copy instead of GitHub, avoiding rate limits and outages Source: thenextweb. Entire plans to open-source its Git backend and benchmark suite.

Why it matters

Coding agents are overloading centralized platforms like GitHub, causing failures. Entire's decentralized approach distributes load and improves speed. Dohmke argues Git was always meant to be distributed Source: thenextweb. The platform also includes a semantic memory layer that logs agent sessions, enabling features like Entire Blame (showing which agent session changed a line) and Entire Review (multi-agent code review).

Performance claims

Entire benchmarks show 570,000 clones per hour and 586 pushes per second from a single repository, though these are not independently verified Source: thenextweb. The company claims up to 25x ahead of competitors like Cursor Origin Source: zdnet. ZDNET notes this addresses GitHub's reliability issues but lacks independent validation.