Merge pull request #2 from DongDongBear/dongdong/feat-wechat-skill-3.30

Dongdong/feat wechat skill 3.30

Author: DophinL

skills/youmind-blog-cover/references/long-running-tasks.md

@@ -16,7 +16,7 @@ When a task involves polling or is expected to take more than a few seconds, **a
 
 ### What to Tell the User
 
-After saving the content to YouMind and providing the link, tell the user (in their language):
+After saving the material to YouMind and providing the link, tell the user (in their language):
 
 > "This may take 10-20 seconds. Would you like me to run this in the background so you can keep chatting? I'll notify you when it's done."
 
@@ -30,7 +30,7 @@ Or, if the agent platform supports subagents (OpenClaw, Claude Code, etc.), the
 Never make the user wait for the entire batch to finish. Deliver results as they become available:
 
 1. **Immediately after saving**: Show the YouMind material link (user can already view the video in YouMind)
-2. **As each result completes**: Output that transcript right away
+2. **As each transcript completes**: Output that transcript right away
 3. **After all done**: Show the summary table
 
 ### Implementation Note for Skill Authors

skills/youmind-deep-research/references/long-running-tasks.md

@@ -16,7 +16,7 @@ When a task involves polling or is expected to take more than a few seconds, **a
 
 ### What to Tell the User
 
-After saving the content to YouMind and providing the link, tell the user (in their language):
+After saving the material to YouMind and providing the link, tell the user (in their language):
 
 > "This may take 10-20 seconds. Would you like me to run this in the background so you can keep chatting? I'll notify you when it's done."
 
@@ -30,7 +30,7 @@ Or, if the agent platform supports subagents (OpenClaw, Claude Code, etc.), the
 Never make the user wait for the entire batch to finish. Deliver results as they become available:
 
 1. **Immediately after saving**: Show the YouMind material link (user can already view the video in YouMind)
-2. **As each result completes**: Output that transcript right away
+2. **As each transcript completes**: Output that transcript right away
 3. **After all done**: Show the summary table
 
 ### Implementation Note for Skill Authors

skills/youmind-image-generator/references/long-running-tasks.md

@@ -16,7 +16,7 @@ When a task involves polling or is expected to take more than a few seconds, **a
 
 ### What to Tell the User
 
-After saving the content to YouMind and providing the link, tell the user (in their language):
+After saving the material to YouMind and providing the link, tell the user (in their language):
 
 > "This may take 10-20 seconds. Would you like me to run this in the background so you can keep chatting? I'll notify you when it's done."
 
@@ -30,7 +30,7 @@ Or, if the agent platform supports subagents (OpenClaw, Claude Code, etc.), the
 Never make the user wait for the entire batch to finish. Deliver results as they become available:
 
 1. **Immediately after saving**: Show the YouMind material link (user can already view the video in YouMind)
-2. **As each result completes**: Output that transcript right away
+2. **As each transcript completes**: Output that transcript right away
 3. **After all done**: Show the summary table
 
 ### Implementation Note for Skill Authors

skills/youmind-slides-generator/references/long-running-tasks.md

@@ -16,7 +16,7 @@ When a task involves polling or is expected to take more than a few seconds, **a
 
 ### What to Tell the User
 
-After saving the content to YouMind and providing the link, tell the user (in their language):
+After saving the material to YouMind and providing the link, tell the user (in their language):
 
 > "This may take 10-20 seconds. Would you like me to run this in the background so you can keep chatting? I'll notify you when it's done."
 
@@ -30,7 +30,7 @@ Or, if the agent platform supports subagents (OpenClaw, Claude Code, etc.), the
 Never make the user wait for the entire batch to finish. Deliver results as they become available:
 
 1. **Immediately after saving**: Show the YouMind material link (user can already view the video in YouMind)
-2. **As each result completes**: Output that transcript right away
+2. **As each transcript completes**: Output that transcript right away
 3. **After all done**: Show the summary table
 
 ### Implementation Note for Skill Authors

skills/youmind-web-clipper/references/long-running-tasks.md

@@ -0,0 +1,41 @@
+# Long-Running Tasks
+
+## Problem
+
+YouMind API calls that involve processing (transcript extraction, image generation, document parsing, etc.) require polling and can take 10-60+ seconds. During this time, the user's main session is blocked — they can't do anything else.
+
+## Solution: Subagent Pattern
+
+When a task involves polling or is expected to take more than a few seconds, **always suggest running it in the background** so the main session stays responsive.
+
+### When to Suggest Background Processing
+
+- Any step that involves polling (checking status in a loop)
+- Batch operations (multiple items to process)
+- Any task the agent estimates will take > 10 seconds
+
+### What to Tell the User
+
+After saving the material to YouMind and providing the link, tell the user (in their language):
+
+> "This may take 10-20 seconds. Would you like me to run this in the background so you can keep chatting? I'll notify you when it's done."
+
+Or, if the agent platform supports subagents (OpenClaw, Claude Code, etc.), the agent should:
+1. Proactively spawn a subagent for the polling + extraction work
+2. Immediately return control to the user
+3. The subagent notifies the user when results are ready
+
+### Progressive Results
+
+Never make the user wait for the entire batch to finish. Deliver results as they become available:
+
+1. **Immediately after saving**: Show the YouMind material link (user can already view the video in YouMind)
+2. **As each transcript completes**: Output that transcript right away
+3. **After all done**: Show the summary table
+
+### Implementation Note for Skill Authors
+
+In your SKILL.md workflow, structure the steps so that:
+- The "save to YouMind" step returns a link immediately
+- The "poll for result" step is clearly separated and marked as the long-running part
+- The agent can naturally split these into "respond now" and "background work"

skills/youmind-webpage-generator/references/long-running-tasks.md

@@ -16,7 +16,7 @@ When a task involves polling or is expected to take more than a few seconds, **a
 
 ### What to Tell the User
 
-After saving the content to YouMind and providing the link, tell the user (in their language):
+After saving the material to YouMind and providing the link, tell the user (in their language):
 
 > "This may take 10-20 seconds. Would you like me to run this in the background so you can keep chatting? I'll notify you when it's done."
 
@@ -30,7 +30,7 @@ Or, if the agent platform supports subagents (OpenClaw, Claude Code, etc.), the
 Never make the user wait for the entire batch to finish. Deliver results as they become available:
 
 1. **Immediately after saving**: Show the YouMind material link (user can already view the video in YouMind)
-2. **As each result completes**: Output that transcript right away
+2. **As each transcript completes**: Output that transcript right away
 3. **After all done**: Show the summary table
 
 ### Implementation Note for Skill Authors

skills/youmind-wechat-article/.clawhubignore

@@ -0,0 +1,5 @@
+references/environment.md
+toolkit/package-lock.json
+toolkit/node_modules/
+image/
+cover/

skills/youmind-wechat-article/.gitignore

@@ -0,0 +1,10 @@
+node_modules/
+dist/
+*.html
+config.yaml
+output/
+__pycache__/
+*.pyc
+.env
+clients/
+toolkit/package-lock.json

skills/youmind-wechat-article/README.md

@@ -0,0 +1,177 @@
+# YouMind WeChat Skill
+
+微信公众号 AI Skill。对 Agent 说一句话，自动跑完选题、写作、配图、排版、发布到草稿箱。
+
+---
+
+## 一句话能干嘛
+
+| 你说 | Skill 做 |
+|------|----------|
+| `给 demo 写一篇公众号文章` | 全自动 8 步：热点 → 选题 → 写作 → SEO → 配图 → 排版 → 发到草稿箱 |
+| `写一篇关于高考志愿的文章` | 跳过热点，直接围绕指定主题走流程 |
+| `把这篇 Markdown 发到草稿箱` | 跳过写作，直接排版发布 |
+| `用紫色 decoration 主题预览` | 换主题换色，即时预览 |
+| `看看最近 7 天文章表现` | 拉数据、分析 top/flop、给下一篇建议 |
+| `根据我的修改学习风格` | 从你的人工改稿中提取经验，下次写得更像你 |
+| `创建新客户 my-brand` | 自动建目录、引导填品牌配置 |
+
+---
+
+## 安装
+
+> 环境要求：Node.js ≥ 18、Python ≥ 3.9、已认证微信公众号（需 API 权限）
+
+```bash
+# 1. 安装依赖
+cd toolkit && npm install && npm run build && cd ..
+pip install -r requirements.txt
+
+# 2. 生成配置文件（如果 config.yaml 不存在）
+cp config.example.yaml config.yaml
+
+# 3. 获取公网 IP（填入微信 IP 白名单，否则无法发布，详见下方「获取本机公网 IP」）
+curl -s https://ifconfig.me
+```
+
+`config.yaml` 需要填写以下凭证：
+
+| 字段 | 必填 | 说明 |
+|------|------|------|
+| `wechat.appid` | **是** | 微信公众号 AppID |
+| `wechat.secret` | **是** | 微信公众号 AppSecret |
+| `wechat.author` | 否 | 文章作者名，默认 "YouMind" |
+| `youmind.api_key` | 推荐 | 用于知识库搜索、联网搜索、文章归档、AI 生图 → [获取 API Key](https://youmind.com/settings/api-keys?utm_source=youmind-wechat-article) |
+| `image.providers.*.api_key` | 否 | 配了哪个就启用哪个（youmind / gemini / openai / doubao） |
+
+### 获取 AppID / AppSecret / 配置 IP 白名单
+
+> 微信开发者平台控制台：<https://developers.weixin.qq.com/platform?tab1=basicInfo&tab2=dev>
+
+**第 1 步 — 进入微信开发者平台**
+
+打开 [微信开发者平台](https://developers.weixin.qq.com/platform?tab1=basicInfo&tab2=dev)，点击首页的 **「前往使用」** 按钮登录。
+
+![步骤1：点击前往使用](image/1.png)
+
+**第 2 步 — 选择公众号**
+
+在「我的业务」面板中，找到并点击 **「公众号」** 进入公众号管理页。
+
+![步骤2：点击公众号](image/2.png)
+
+**第 3 步 — 复制 AppID、AppSecret 并配置 IP 白名单**
+
+在公众号 → **基础信息** 页面：
+
+1. **AppID** — 顶部「基础信息」区域直接复制，填入 `config.yaml` 的 `wechat.appid`
+2. **AppSecret** — 「开发密钥」区域，点击 **重置** 获取（只展示一次，请立即保存），填入 `wechat.secret`
+3. **API IP 白名单** — 同一区域，点击 **编辑**，将你的公网 IP 粘贴进去
+
+![步骤3：AppID、AppSecret 和 IP 白名单位置](image/3.png)
+
+### 获取本机公网 IP
+
+家庭宽带 IP 会变，发布报 IP 错误时重新获取并更新白名单即可。
+
+**macOS**
+
+```bash
+curl -s https://httpbin.org/ip | python3 -c "import sys,json; print(json.load(sys.stdin)['origin'])"
+# 或者
+curl -s https://ifconfig.me
+```
+
+**Windows（PowerShell）**
+
+```powershell
+(Invoke-WebRequest -Uri "https://ifconfig.me" -UseBasicParsing).Content.Trim()
+# 或者
+(Invoke-RestMethod -Uri "https://httpbin.org/ip").origin
+```
+
+**Linux**
+
+```bash
+curl -s https://ifconfig.me
+# 或者
+curl -s https://httpbin.org/ip | python3 -c "import sys,json; print(json.load(sys.stdin)['origin'])"
+```
+
+> **提示**：拿到 IP 后，回到上面第 3 步的「API IP 白名单」→ 编辑，粘贴保存即可。
+
+---
+
+## 使用技巧
+
+### 两种运行模式
+
+- **自动模式**（默认）：全程自动跑，只在生成配图前问一次图片风格偏好
+- **交互模式**：说"让我来选题" / "交互模式"，会在选题、框架、配图、主题环节暂停让你选
+
+### 主题系统
+
+4 款内置主题，搭配任意 HEX 色值：
+
+| 主题 | 风格 | 适合 |
+|------|------|------|
+| `simple` | 简约干净 | 日常推送、知识科普 |
+| `center` | 居中排版 | 短篇、金句、情感 |
+| `decoration` | 装饰线条 | 品牌感强的内容 |
+| `prominent` | 大标题 | 深度长文、观点输出 |
+
+<!-- TODO: 主题对比截图 -->
+
+想要更深的定制？用 Theme DSL 写一个自定义主题 JSON，放到 `clients/{client}/themes/` 下，发布时加 `--custom-theme` 即可。
+
+### 配图降级链
+
+Skill 按以下顺序尝试生成配图，任何一环成功就继续，全挂也不中断流程：
+
+```
+AI 生图（你配的 provider）→ Nano Banana Pro 图库搜索 → CDN 预制封面下载 → 只输出 prompt
+```
+
+### 多客户管理
+
+每个客户一个目录，互不干扰：
+
+```
+clients/demo/
+├── style.yaml      # 品牌调性、目标读者、禁用词
+├── playbook.md     # 写作手册（自动生成或手写）
+├── history.yaml    # 已发布记录（去重用）
+├── corpus/         # 历史语料
+├── lessons/        # 人工改稿经验
+└── themes/         # 专属主题
+```
+
+对 Agent 说 `创建新客户 xxx` 即可自动初始化。
+
+### 让文章越写越像你
+
+1. **喂语料**：往 `clients/{client}/corpus/` 里放 20+ 篇你的历史文章，跑 `build-playbook` 生成写作手册
+2. **改稿学习**：发布后手动改，然后说"根据我的修改学习风格"——Skill 会提取差异存到 `lessons/`
+3. **手册迭代**：每积累 5 条经验，自动刷新 `playbook.md`
+
+### 流程不会断
+
+每一步都有 fallback。热点抓不到就联网搜，联网搜也挂就问你；图生不出来就搜图库；发布失败就生成本地 HTML。单步失败只会跳过并标注，不会卡死。
+
+---
+
+## 常见问题
+
+**发布报 IP 错误** — 公网 IP 变了。重跑 `curl -s https://ifconfig.me` 拿新 IP，更新微信白名单（详见上方「获取本机公网 IP」章节）。
+
+**图片生成失败** — 不影响发布。Skill 会自动走降级链。想用特定 provider 就在 `config.yaml` 里填对应 key。
+
+**文章有 AI 味** — 在 `style.yaml` 里写清楚你的调性；多喂历史语料建 playbook；发布后改稿再跑"学习风格"。三管齐下效果最好。
+
+**怎么自定义排版？** — 三个层级：① 对话里指定颜色字号 → ② 写 Theme DSL JSON → ③ 搭配设计类 Skill 深度定制。
+
+---
+
+## 许可证
+
+MIT