Skip to content

Releases: YogurtJ/wechat-devtools-automator

v0.2.0-beta.4 — MIT license and README polish

24 Mar 18:06
@YogurtJ YogurtJ
v0.2.0-beta.4
aefd34d

Choose a tag to compare

View all tags
v0.2.0-beta.4 — MIT license and README polish Pre-release
Pre-release

Summary

wechat-devtools-automator is now MIT-licensed.

This beta repository gives agents a real visual loop inside WeChat DevTools for 微信小程序 work: route discovery, query-based opening, interaction replay, scrolling, screenshots, and console / exception evidence in report.json.

What's new in beta.4

  • Switch the repository license to MIT
  • Replace the temporary license notice with a real LICENSE
  • Add README badges and sharpen the first-screen positioning
  • Keep the repo bilingual with README.md and README.zh-CN.md
  • Preserve the broader agent-tool positioning while keeping Codex .skill installation as an option

How to test

git clone https://github.com/YogurtJ/wechat-devtools-automator.git
cd wechat-devtools-automator
export WDA="$(pwd)/scripts/wechat_devtools_automator.sh"
"$WDA" doctor --project /path/to/mini-program-project
"$WDA" routes --project /path/to/mini-program-project
"$WDA" shot --project /path/to/mini-program-project --route "<copy-a-route-from-routes-output>"

Artifacts land under <project>/output/wechat-devtools-automator/<run-id>/.

Included asset

  • wechat-devtools-automator.skill

Compatibility

  • Validated on macOS 15.5 / Node v24.13.1 / WeChat DevTools 2.01.2510280

License

  • MIT

Docs

Loading

v0.2.0-beta.3 — bilingual README and broader agent support

24 Mar 17:21
@YogurtJ YogurtJ
v0.2.0-beta.3
7ca3c8b

Choose a tag to compare

View all tags
v0.2.0-beta.3 — bilingual README and broader agent support Pre-release
Pre-release

Summary

wechat-devtools-automator is a public GitHub beta for agentic WeChat Mini Program visual verification inside WeChat DevTools. It helps agents open real routes, interact with the UI, scroll below the fold, capture screenshots, and collect console / exception output into a rerunnable evidence folder.

The packaged .skill targets Codex today, while the same wrapper and scripts can also be called from Claude Code, Cursor, OpenCode, OpenClaw, or similar agent tools that can run local shell commands.

What's new in beta.3

  • Remove inline README screenshots and keep the landing page cleaner and text-first
  • Add README.zh-CN.md so Chinese developers have a first-class onboarding path
  • Reposition the repo as agent-tool-friendly instead of over-framing it as Codex-only
  • Add a generic wrapper-based usage path for non-Codex agent tools

How to test

git clone https://github.com/YogurtJ/wechat-devtools-automator.git
cd wechat-devtools-automator
export WDA="$(pwd)/scripts/wechat_devtools_automator.sh"
"$WDA" doctor --project /path/to/mini-program-project
"$WDA" routes --project /path/to/mini-program-project
"$WDA" shot --project /path/to/mini-program-project --route "<copy-a-route-from-routes-output>"

Artifacts land under <project>/output/wechat-devtools-automator/<run-id>/.

Included asset

  • wechat-devtools-automator.skill

Compatibility

  • Validated on macOS 15.5 / Node v24.13.1 / WeChat DevTools 2.01.2510280

Caveats

  • Requires a local WeChat DevTools installation and a real Mini Program project
  • Visual results can vary by OS, fonts, zoom, and mock / API data
  • Current license posture is UNLICENSED; this public beta is for evaluation and feedback, not an open-source license grant

Docs

Loading

v0.2.0-beta.2 — WeChat DevTools Automator beta

24 Mar 17:13
@YogurtJ YogurtJ
v0.2.0-beta.2
a8b90a8

Choose a tag to compare

View all tags
v0.2.0-beta.2 — WeChat DevTools Automator beta Pre-release
Pre-release

Summary

wechat-devtools-automator is a beta Codex skill that gives agents real visual evidence inside WeChat DevTools for 微信小程序 work. It can discover routes, open pages with query params, perform interactions, scroll below the fold, capture screenshots, and collect console / exception output into a rerunnable artifact folder.

What's new

  • Route-aware CLI flow for doctor, routes, shot, scroll-shot, and gui-shot
  • Interaction helpers for --tap, --longpress, --input, --trigger, and --action
  • Evidence bundle per run with screenshots plus report.json
  • Packaging hardening so the .skill artifact excludes nested dist/, .git/, and scratch files
  • Demo gallery, install guide, release checklist, and trigger QA docs for easier onboarding

How to test

export CODEX_HOME="${CODEX_HOME:-$HOME/.codex}"
git clone https://github.com/YogurtJ/wechat-devtools-automator.git \
  "$CODEX_HOME/skills/wechat-devtools-automator"
export WDA="$CODEX_HOME/skills/wechat-devtools-automator/scripts/wechat_devtools_automator.sh"
"$WDA" doctor --project "$(pwd)"
"$WDA" routes --project "$(pwd)"
"$WDA" shot --project "$(pwd)" --route "<copy-a-route-from-routes-output>"

Artifacts land under <project>/output/wechat-devtools-automator/<run-id>/.

Included asset

  • wechat-devtools-automator.skill

Compatibility

  • Validated on macOS 15.5 / Node v24.13.1 / WeChat DevTools 2.01.2510280

Caveats

  • Requires a local WeChat DevTools installation and a real Mini Program project
  • Visual results can vary by OS, fonts, zoom, and mock / API data
  • Current license posture is UNLICENSED; this GitHub beta is for evaluation and feedback, not an open-source license grant

Links

Loading