一个基于 UniApp 的微信小程序,通过 AI 图像识别帮助用户诊断植物问题,提供专业的养护建议。
目标用户: 阳台种菜人士、办公室绿植爱好者
核心痛点: 绿植生病无法马上得到可靠的解决方案
解决方案: 拍照 + 混元 Vision AI 诊断,提供专业养护建议
变现模式: 订阅制会员
- 框架: UniApp 3.0 + Vue 3
- 状态管理: Pinia
- 样式框架: Tailwind CSS
- 构建工具: Vite
- 云服务: 腾讯云 CloudBase
- AI 能力: Qwen/混元 Vision 模型
微信小程序本地开发可以直连本机 CloudBase HTTP 云函数,无需先部署到云端:
npm run dev:functions:install
npm run dev:functions
npm run dev:mp-weixin:local-functions用微信开发者工具打开 dist/dev/mp-weixin。如需真机或局域网访问,可使用:
npm run dev:mp-weixin:local-functions:lan未设置 VITE_API_BASE_URL 时仍走原有 CloudBase gateway;线上/生产构建不允许使用本地或非 HTTPS 的 API Base URL。完整说明见 docs/local-cloudbase-functions-debugging.md。
This package contains a complete repository-ready Codex configuration:
AGENTS.md: concise repository-level entrypoint..codex/config.toml: subagent runtime limits..codex/agents/*.toml: task planner, code explorer, architecture reviewer, implementer, QA reviewer, docs keeper, release ops.docs/ai-rules/*.md: split rule files for project hard rules, language policy, CloudBase deployment, replay, subagent workflow, and the full original CloudBase guide.docs/ai-tasks/: place task notes here.docs/ai-runs/: place subagent handoff records here.
Copy all files into your repository root.
git add AGENTS.md .codex docs/ai-rules docs/ai-tasks docs/ai-runs docs/adr
git commit -m "chore: add codex ai team rules and subagents"The original uploaded AGENTS.md is preserved as:
docs/ai-rules/AGENTS.original.md