This article tackles the challenge of selecting AI coding tools for Chinese companies, comparing three options: Kiro, Trae, and Qoder. The author proposes a practical testing approach: having each tool create a simple Android app, then compiling and uploading it via GitHub Action workflows to assess their viability as development assistants. Test results indicate Kiro is functional, Trae underperformed, while Qoder remains untested. This workflow-based testing methodology offers an objective evaluation framework for businesses choosing AI programming tools, enabling more informed decision-making.
