WeChat AI Integration Advisor

Problem

Mini program teams integrating WeChat AI often need to decide between development mode and automatic mode, and to expose products, orders, reservations, payments, delivery, and query flows in a shape that agents can understand, call, and recommend. This skill turns business context into an implementation plan instead of only explaining the concept.

How It Works

• Mode diagnosis: evaluates industry, mini program status, development resources, and available APIs to recommend a mode.
• Capability breakdown: generates atomic API lists with names, parameters, returns, prerequisites, and flows such as search, recommendation, order, payment, and status checks.
• Templates and code: drafts AGENTS.md, business SKILL.md, and mcp.json, then optionally uses templates/ to produce app.json, apis/, components/, and page-meta.json skeletons.
• GEO structuring: adds product, store, service, price, inventory, delivery, promotion, and scenario tags so AI can more easily understand and invoke capabilities, not traditional SEO.
• Launch checks: provides test matrices, release checklists, fallback behavior, compliance risks, and mcp.json best-practice checks.

Boundaries

It is most useful for mini programs with products, orders, reservations, payments, delivery, or similar transaction flows, and for teams using automatic mode for pages, knowledge bases, and metadata as a fallback. It does not promise rankings, traffic, or recommendations, does not recommend automatic mode alone for complex transactions, and does not provide ways to bypass WeChat review, payment confirmation, privacy authorization, or content safety rules. Mini program AI is in beta and code submission for review is not currently open.