Mini Program Engineering Skill Suite
Nine skills for WeChat, Alipay and Douyin mini programs: intake, architecture, debugging, verification and release readiness
- 97 ★ on the repository
- Last commit
- MIT
Minimal access
Why this level
- The suite is instructions and check scripts, with external actions left to the user
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/mini-program-engineering-skill-suiteDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewMinimal accessThe tool runs nothing and makes no network calls.
- The suite is instructions and check scripts, with external actions left to the user
This is third-party code. Review the repository files before installing.
What it does
The suite guides an agent through the full mini program lifecycle, from requirement intake to release readiness. The root skill is invoked explicitly, and nine specialist skills sit inside: `mini-program-project-intake-skill` (intake and scope), `mini-program-product-spec-skill`, `mini-program-architecture-skill`, `mini-program-implementation-skill`, `mini-program-ui-device-skill`, `mini-program-debugging-skill`, `mini-program-verification-skill`, `mini-program-release-skill` and `wechat-mini-program-platform-skill`. The `foundation/` layer sets the discipline: every status claim is backed by evidence or labeled unknown. Project diagnosis detects the stack (Taro, uni-app or native) and the target platform, and a shared fact layer collects the rules of WeChat, Alipay and Douyin. Rule freshness is checked against official docs at run time. The suite does not install project dependencies, create cloud resources, upload packages or submit for review: external actions stay with the user and need explicit permission. There is an `install.sh` installer with targets for Claude Code, Codex, GitHub Copilot and Cursor, and an install-by-request route where you give the agent the repository link.
Who it is for. Mini program developers on Taro, uni-app or a native stack who hand an agent project review, debugging and platform review preparation.
Good fit when
- An existing mini program project is taken over and its stack, platform and scope must be understood first
- A mini program is prepared for WeChat review and needs verification evidence and a list of manual steps
- A form loses input after an error and recovery has to be checked on a device
- One codebase ships to WeChat, Alipay and Douyin and platform rules differ
Not a fit when
- The project is not about mini programs: the suite targets WeChat, Alipay and Douyin
- You want the agent to upload the package and submit it for review: the suite leaves external actions to the user
- You need only short tips for one framework: the suite is large and works in stages
Example request
Use mini-program-engineering-skill-suite: review the mini program project, detect the platform and stack and draft a pre-release checklistLimitations
- Alipay and Douyin rules are checked against official documentation at run time, and deterministic fingerprint monitoring exists only for WeChat, as the README says.
- The suite's strength, evidence discipline, slows simple tasks: every status claim needs confirmations.
- The installer does not overwrite existing folders without `--force`, and with it makes a backup. The old copy should be moved out of skill discovery directories.
- Python 3.9 or newer and network access for checking current rules are required.
How to disable. Delete the installed suite folder from the agent's skills directory. With `--force` the installer leaves a backup beside it, which should also be moved out of skill discovery directories.
FAQ
How do I install the mini program suite?
Clone the repository and run `bash install.sh --target auto`. For project-level Copilot or Cursor install use `--target copilot` or `--target cursor` with `--project` and the project path. You can also simply ask an agent: give the repository link and say to install the skill.
Which platforms are supported?
WeChat, Alipay and Douyin, on Taro, uni-app and native stacks. Diagnosis detects the project's stack and target platform.
What does the suite do on its own and what does it leave to me?
It prepares evidence and hints and does not touch dependencies, cloud resources, package upload, review submission or release. The user performs those steps with explicit permission.
Where do platform rules come from?
From the shared fact layer `facts.md` and from official documentation the agent checks at run time. Weekly drift checks propose updates as PRs or issues.
How do I update an installed copy?
Download the new package from Releases and run `bash install.sh --target auto --force`. The installer first moves the old folder into a timestamped backup.
Discussion
No questions or reviews yet.
Related
SkillsA playbook for AI assistants on WeChat mini programs: from picking a topic and registering to building, monetizing and promoting.
A skill with scripts for building iOS apps and driving the simulator through the accessibility tree, with compact output for agents
A set of skills for iOS 26+, Swift 6.3 and SwiftUI that gives the agent knowledge of modern Apple frameworks
Official set of 30 skills from the Android team: Jetpack Compose, AGP 9 upgrade, R8, Perfetto profiling, Navigation 3, CameraX, Google Play policies
wechat-devtools-mcp: MCP for WeChat DevTools
wechat-devtools-mcp
An MCP server and skill for WeChat DevTools: an agent compiles, previews, debugs and auto-tests WeChat Mini Programs
A Claude Code skill that builds Expo mobile apps with onboarding, paid subscriptions, AdMob ads, analytics and localization.
README in short
The README states the suite's purpose, the main ideas of versions 3.6.0 and 3.7.0, a feature summary (a three-platform fact layer, rule freshness, evaluation in real agent sessions, evidence discipline), a component map and principles. The practical part covers installing with the installer or by asking an agent, invocation, staged work, updating, limits and package integrity.