H

feat: AI responses always in English

feat: lenient AI evaluation with clear type detection and improved feedback UI

feat: AI evaluation for all challenge types including terminal/git

fix(mistral): don't include correct query in SQL suggestions

fix(challenges): switch SQL evaluation to AI-only with strict prompt

H

H

H

H

H

H

H

H

H

H

H

H

H

H

H

feat: add backend Docker setup for HF Spaces