Agentic Coding Durability Evaluation Set
This repository collects real-world coding tasks across multiple languages and tech stacks, with varying difficulty. It helps developers evaluate how durable different agentic coding products are when tackling diverse, practical programming work.
- Choose a repository for the language/stack you care about.
- Open the project folder and read Prompt.md or Prompt.zh.md.
- Run the prompt in your chosen AI coding product (e.g., Qoder, Cursor, Windsurf, Kiro, Claude Code).
- Iterate until you personally judge the task “done.”
- This evaluation set intentionally does not include automated unit/integration/UI tests to decide completion.
- We rely on your human judgment, similar to real-world work where completeness, test coverage, visual acceptance, and maintainability vary by person.
- Keep the model tier comparable across products whenever possible.
- Record and compare consumption within your paid plan to evaluate durability.
🌐 JavaScript/TypeScript Projects
- vite-react-ts-demo: React.js demo with Vite.js, TypeScript and Tailwind CSS
- react-ts-demo: React.js demo with TypeScript
- react-js-demo: React.js demo with JavaScript
- nextjs-demo: A React full-stack application demo project based on Next.js 16
- hono-node-demo: A Node.js Web application demo project based on the Hono framework
- fresh-demo: A Deno full-stack Web application demo project based on the Fresh framework
- bun-react-demo: A full-stack application demo project based on Bun and React
- gitea: Forked from gitea
- gin-demo: A Go Web application demo project based on the Gin framework