A compact software-engineering evaluation benchmark. It contains five realistic Python engineering tasks, each with a verified reference solution, and a small CLI that applies a candidate patch (human ...
New here? Follow chapter 0 first. For full acceptance, run chapters 1–5 first. Image builds, SDKs, security databases and hardware tests take substantially longer than UI/navigation tests. Chapter 11 ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results