An OpenAI Codex skill for shipping WeChat Mini Program projects that use Tencent CloudBase. It captures a practical workflow for implementation, CloudBase debugging, admin-to-mini-program integration testing, customer trial handoff, clean ZIP packaging, and optional WeChat file delivery.
- WeChat Mini Program frontend implementation and debugging
- Tencent CloudBase setup checks, including
envId, AppID association, NoSQL collections, permissions, cloud functions, storage, and static hosting - Admin backend to mini program data synchronization
- Guided intake for no-code or incomplete projects
- WeChat Developer Tools error triage
- Full customer trial flow from backend product upload to mini program display/order testing
- Clean source-code packaging for delivery
- File handoff preparation for sending a ZIP to a WeChat contact or group
Clone this repository into your Codex skills directory:
mkdir -p ~/.codex/skills
git clone https://github.com/OwenZhao9/tencent-cloudbase-wechat-miniprogram-delivery.git \
~/.codex/skills/tencent-cloudbase-wechat-miniprogram-deliveryRestart Codex or start a new session so the skill metadata is discovered.
Use prompts like:
Use Tencent CloudBase WeChat Mini Program Delivery. I have no code yet. Ask me the required setup questions first, then build the MVP.
Use Tencent CloudBase WeChat Mini Program Delivery to debug this mini program and admin backend.
Use this skill to test the full flow: admin creates a product, mini program shows it, user places an order, admin processes it.
Use this skill to package the current WeChat Mini Program and CloudBase admin project for customer delivery.
SKILL.md
agents/openai.yaml
references/
cloudbase-miniprogram-debugging.md
delivery-checklist.md
intake-questionnaire.md
product-design-qa.md
scripts/
package_delivery.sh
For projects with no code or incomplete requirements, the skill asks only the blocking questions first:
- industry and transaction type
- required MVP features
- Mini Program AppID and CloudBase
envIdavailability - customer trial vs source-code delivery
- brand assets and initial product data
If you want speed, tell Codex to choose defaults. The skill will proceed with an MVP mall flow and list assumptions before implementation.
The packaging script creates a clean ZIP that excludes dependencies, Git metadata, local IDE state, AI working folders, private environment files, and WeChat private project config.
./scripts/package_delivery.sh \
--admin /path/to/admin-project \
--miniprogram /path/to/miniprogram-root \
--out /path/to/output \
--name customer-projectUse --include-build only when you intentionally want dist/ or build/ artifacts included.
The skill intentionally prioritizes root-cause fixes over fake fallback data:
- Do not mask CloudBase permission or collection errors with mock data unless explicitly requested.
- Verify AppID,
envId,wx.cloud.init, collection names, permissions, and cloud functions first. - Treat WeChat Developer Tools internal warnings differently from real application errors.
- Prove admin-to-mini-program sync with a real mutation before handoff.
MIT