Ex-GitHub CEO Launches Entire: The Git Forge Built Entirely for AI Agents

According to ZDNET reporting, former GitHub CEO Thomas Dohmke has launched Entire, a new developer platform engineered entirely around the reality of AI-driven coding. As coding agents generate a massive volume of code, traditional git workflows struggle with infrastructure bottlenecks and review fatigue. Entire introduces a decentralized Git network to prevent the kinds of outages that plague legacy platforms when AI traffic spikes.

How Entire Replaces Traditional Pull Requests

Instead of reviewing massive text diffs line by line, Entire ditches standard pull requests for a concept called Trails. According to ZDNET, Trails act like an airplane pilot's heads-up display, showing customized evaluations scored automatically by agents on every code push. Developers can track token spend per trail and let agents handle fixes in a loop until all evaluations pass.

Once checks clear, developers can merge changes without manually reading every single line of code. The platform also features Continuous Memory, keeping past session logs directly in the repository. This allows agents to reference past context, preventing repeated mistakes and cutting token consumption by up to 71%.

How it works in practice

Entire is available for developers to test today. Here is how teams can start using the platform to manage agent-driven workflows:

• Create or migrate repositories: Set up private code repositories across decentralized global locations, starting with the US, India, Australia, and the EU.
• Use the Entire CLI: Access the open-source command-line interface to let coding agents interact with codebase data directly in a loop.
• Monitor Trails: Replace pull request reviews with agent-scored evals that automatically verify the success of code changes.
• Merge without line-by-line diffs: Authorize code merges once automated evaluations are successfully satisfied.

Is it worth changing your workflow?

For development teams overwhelmed by the volume of AI-generated code, Entire offers a compelling infrastructure upgrade. ZDNET points out that the platform provides hosting speeds up to 89x faster than GitHub and six times faster than Cursor origin. While migrating an existing codebase requires effort, the reduction in token waste and the elimination of manual pull request bottlenecks make it a strong option for modern engineering teams.