WeChat Mini Program Vibe Coding Starter

Problem

Non-engineers can describe a product idea but often get stuck on WeChat mini program structure, state, sharing, cloud development, and real-device validation. Asking an AI for a full build can lead to over-engineering, broad edits, and changes that cannot be verified.

How it works

• Translate a vague request into a product path: user, core scenario, and shortest loop, such as home page, submit, and view result.
• Prefer lightweight choices first: pages with WXML, WXSS, and JS; local storage for personal state; WeChat Cloud Development only when data needs to cross users or devices.
• Work in atomic tasks: before editing, list affected files, risks, and acceptance checks; after editing, summarize changes, self-tests, device paths, and rollback points.
• Flag high-risk features such as sharing, permissions, posters, location, and cloud functions; check onShareAppMessage, avoid unnecessary permissions, and redeploy cloud functions before testing.
• Compare options in tables and estimate work as XS/S/M/L to avoid large diffs.

Boundaries

It fits early mini program MVPs, beginner learning, and small validation loops. It is not intended for large backends, complex permission systems, high-concurrency payments, or production-grade architecture. Review, privacy, and real data still require WeChat rules and actual business validation.