SWE-Universe is a factory-like system that turns real GitHub pull requests into safe, repeatable coding practice worlds with automatic checkers.
This paper teaches code AIs to work more like real software engineers by training them in the middle of their learning using real development workflows.
Coding agents waste most of their tokens just reading giant files, which makes them slow and expensive.
The paper introduces Agentic Rubrics, a new way to check code fixes without running the code by creating a smart checklist from the project itself.
Coding agents used to fix software rely on feedback; unit tests give only pass/fail signals that are often noisy or missing.