HexaShell coding-agent harness sets a new RoboDojo score
With Codex and GPT-6-Astra, success is 39.7% on 42 tasks; tight insertion is 8.3% while pushing reaches 92.5%.
3 stories, newest first · one card per story
With Codex and GPT-6-Astra, success is 39.7% on 42 tasks; tight insertion is 8.3% while pushing reaches 92.5%.
Code-Only-as-Policy makes robot decisions in code, reaching 70.24% success on 42 RoboDojo tasks.
PhysEvo lets a frozen Astra model recursively improve robot skills, reaching 84% success on real tasks.
Today 2.3K readers · 2.8K views · 408 reads · all time 108K views