Entire launches a Git forge built around coding agents, with live eval scores on every pull request
Pull requests become a running record that agents update on each push, with merges blocked until review findings are closed.
Entire has launched a Git forge built for collaboration between developers and AI coding agents, arguing that the pull request model has become an audit log that only the agent reads. Its central feature, called trails, attaches eval scores to each pull request that refresh on every push, covering code quality, security, slop and whether human review is needed. Review agents run in cloud sandboxes and their findings must be resolved before a merge is permitted, and repository-level gates can set further approval conditions. Entire also introduces continuous memory through two open-source tools, entire-graph and entire-brain, which let agents reuse repository knowledge across sessions. The platform claims pushes up to 89 times faster than GitHub and integrates with CI services such as Buildkite and Depot.
Why it matters
For people who follow software development, the change is in what a pull request is for. Entire argues the traditional pull request has become a record mainly read by agents, and its design puts quality and security checks at the center of the review. Merges can be held until every review finding is resolved or closed, and the claimed push speed of up to 89 times faster than GitHub is an early company claim rather than an independently measured result. Whether teams accept this model will show how far agent-written code changes the way software is reviewed and shipped.
Matters or hype?
Does this story matter, or is it hype? Vote before you see what everyone else said.