SonarSource study finds clean code cuts coding-agent token use, not pass rates

The May 19th arXiv preprint tested Claude Code on paired repositories and found fewer tokens and file revisits on cleaner code.

By · Published

Illustration comparing clean and messy repositories from the SonarSource study, showing the Claude Code agent using fewer tokens and revisiting files less on the clean code.

The May 19th arXiv preprint tested Claude Code on paired repositories and found fewer tokens and file revisits on cleaner code.