The headline is about Ruby being AI-friendly, with nuances. Let me craft a concise headline.

Short summary

Ruby's conventions help AI models write code cleanly, but navigating dependencies across large codebases remains a challenge. A structural-dependency map dramatically improves recall (from 24-40% to 63-97% in tests). The benefit scales inversely with model size, meaning smaller models gain more from structural augmentation than frontier models.

• •Ruby conventions aid code generation but not cross-file dependency navigation
• •Structural maps boost recall from 24-40% to 63-97% across tested repos
• •Smaller models gain more from structural augmentation than frontier models

Generated with AI, which can make mistakes.