AiPexBase BaaS Dev Flow

Problem

In frontend-heavy demos, internal tools, and lightweight business apps, requirements, schemas, SDK setup, and deployment configuration often live in separate manual steps. This can lead to mismatched field types, scattered baseUrl and apiKey values, and accidental data sharing across projects. This skill targets lightweight systems built around data tables, files, AI features, and notification delivery. It organizes the work around AiPexBase BaaS: clarify requirements, generate prd.md, create the app and tables, build the frontend, and prepare deployment.

How It Works

• Fixed workflow: it follows prd.md → app creation → schema design → frontend integration → deployment, with new app and iteration steps automated and deployment gated by confirmation.
• Project isolation: each project uses its own baas-config.json containing baseUrl, apiKey, and appId, reducing cross-project data mixing.
• Schema constraints: app-schema.json generation enforces primary keys, typed columns, explicit VARCHAR(n) lengths, FULLTEXT for long text, and non-required datetime fields, lowering schema/form mismatches.
• Frontend integration: pages consume aipexbase-js through createClient(), using shared templates for HTML, Vue, or React and reusing platform extensions such as AI, notifications, and location features.

Boundaries

It fits frontend-driven apps, data management pages, and lightweight business loops built on AiPexBase BaaS. It is not intended for heavy custom server-side logic, complex microservice topology, or systems that must bypass BaaS authentication. Usage requires a valid baseUrl and manageToken, and the platform’s field semantics, especially isNullable, may differ from conventional database conventions.