# Wechat Lab Builder

> Use this skill for experimental changes that must stay in lab paths only (sandbox/test workspaces) and must never modify production project code.

- Skill: `sdaxiji-beep/wechat-lab-builder` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add sdaxiji-beep/wechat-lab-builder`
- Raw SKILL.md: https://api.skillmd.com/api/skills/sdaxiji-beep/wechat-lab-builder/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: sdaxiji-beep (https://skillmd.com/u/sdaxiji-beep)
- Updated: 2026-09-22
- Page: https://skillmd.com/skills/sdaxiji-beep/wechat-lab-builder

---


# WeChat Lab Builder

Use this skill when the user wants to experiment safely without touching production code.

## Safe paths

- `sandbox/fake-project`
- repo-local scratch workspaces only

## Entry

```powershell
$RepoRoot = (Get-Location).Path
. (Join-Path $RepoRoot "scripts\wechat.ps1")

Invoke-SafeWrite `
  -ProjectPath (Join-Path $RepoRoot "sandbox\fake-project") `
  -Description "lab: <change>" `
  -RequireConfirm $false `
  -WriteAction { <edit action> }
```

```powershell
$RepoRoot = (Get-Location).Path
powershell -ExecutionPolicy Bypass -File (Join-Path $RepoRoot "scripts\test-wechat-skill.ps1") -SkipSmoke -Tag fast
```

## Hard rules

- Never write to production project code in lab mode.
- Use guarded write helpers instead of raw writes.
- Keep deploy actions out of lab requests unless explicitly requested.