Thomas Dohmke debuts Entire with an open source CLI that records the reasoning behind AI-generated code, aiming to solve review bottlenecks and bring auditability to AI-led software development.
Entire has launched Checkpoints, an open source command-line tool designed to make AI-generated code transparent and reviewable, marking the first release from a new developer platform founded by former GitHub CEO Thomas Dohmke.
Checkpoints records the reasoning and instructions behind AI-produced code, capturing original prompts, transcripts, agent decision steps and implementation logic, and storing that context alongside the code itself. The tool supports Anthropic Claude Code and Google Gemini CLI. Entire plans to follow an open source business model, with Checkpoints as the foundation.
The move addresses a growing bottleneck: AI coding agents now generate software faster than teams can audit it. Security and zero-trust policies still require human sign-off, turning review into a productivity constraint. Entire aims to add traceability so teams understand not only what changed, but why.
“We are living through an agent boom, and now massive volumes of code are being generated faster than any human could reasonably understand. The truth is, our manual system of software production, from issues, to git repositories, to pull requests, to deployment, was never designed for the era of AI in the first place,” said Thomas Dohmke, Founder, Entire.
Beyond Checkpoints, Entire is building a three-layer platform comprising a Git-compatible database, a universal semantic reasoning layer, and an AI-native interface to manage fleets of AI agents alongside tools such as GitHub and GitLab.
Dohmke, who led GitHub from 2021 to August 2025 and scaled Copilot, has raised $60 million in seed funding at a $300 million valuation, led by Felicis, with participation from M12, Madrona and other backers.
