Ex-GitHub CEO launches a GitHub rival that reinvents the pull request

The core of Entire revolves around its Git-compatible distributed source control server, which debuted earlier this summer, boasting speeds up to 89x faster than GitHub and 6x faster than Cursor's Origin.

This new venture, founded by former GitHub CEO Thomas Dohmke, launched on Monday as a direct competitor to GitHub, introducing a novel agentic code review system named Trails to replace the conventional pull request model.

Dohmke shared this excitement last week during an interview at the Linux Foundation's AGNTCon + MCPcon in San Jose, California, stating, "It's the moment we've been waiting for." He elaborated that bringing the Git hosting and infrastructure layer together, alongside continuous integration, continuous delivery, continuous memory, and Trails, represents a "reinvention of the pull request."

Entire posits that the current development constraint is not writing code—which agents handle—but rather the code review process. Unlike some rivals, the company believes solving this bottleneck demands an entirely new platform.

Trails functions as Entire's unique approach to code review. Instead of manual developer scrutiny, it provides a dashboard featuring customizable evaluations scored by agents. These evals can gauge things like the risk level of a code change or how much intent drift occurred during an LLM-driven session.

These evaluations execute on every push, with results accessible via the web interface or the open-source CLI. Coding agents can then leverage this feedback to revise code, which the developer subsequently validates. Entire asserts that if the evals return green, merging the code without extensive human review is acceptable.

Beyond merely adding an AI reviewer to a pull request, Entire incorporates the intent behind a change and the agent's work directly into the review process. Developers can use preset review agents or define their own configurations in a JSON file, supporting Claude and OpenAI models.

The platform also features native hosting for private repositories at launch; public repositories and self-hosted Git nodes are slated for future release. Furthermore, Entire's Continuous Memory pillar syncs AI agent sessions into a project repository, preserving rejected approaches and developer feedback for agents to query later.

Dohmke anticipates that agent run costs may become the next hurdle, suggesting that teams might find using services like Entire cheaper than generating replacements using tokens. The platform is free, but customers pay only for the tokens consumed by the review agents. Currently, Entire offers all users a $100 starter credit.