W

微信公众号文章创作

Contentv1.0.0
By user_5dda6a29Updated 2026-10-01wechat-content-studio@user_5dda6a29/wechat-content-studio
Article WritingRewriting

According to AlphaMoat, as of 2026-09, 微信公众号文章创作 has 3,860 downloads and 3 stars — ranked #2,705 of 5,078 AI skills overall, and #187 of 362 in Content.

Installs
402
Downloads
3.9K
Stars
3
Overall Rank
#2,705
Top 54% overall · Top 52% in Content
PlatformInstallsDownloadsStarsComments
SkillHub ↗4023.9K3—

Platforms count differently and update at different times.

What this skill does

It searches by source or supplied links, merges articles and prepares rewrites and covers, with an option to retrieve content without automatic publishing. It depends on Node.js and extraction/layout components; rewriting and covers need their respective API keys. Publishing requires official-account credentials and an allowlisted server IP. Generated content and copyright still need review.

Profile published 2026-10-27

What you can do with it

  • Collect source material and prepare an official-account article and cover

When to use it

  • When building a WeChat content workflow from a topic or links

Example prompts

Generated from the skill’s documentation as examples; not results of actual runs.

Search and merge material without automatic publishing, then list the configuration needed for rewriting, covers and publication.

Monthly Trend

Downloads3.9K
3632.2K4.1K2026-03 · Downloads 6222026-05 · Downloads 1.1K2026-06 · Downloads 1.5K2026-07 · Downloads 2.2K2026-08 · Downloads 2.7K2026-09 · Downloads 3.9K26-0326-09
Stars3
1232026-03 · Stars 12026-05 · Stars 12026-06 · Stars 12026-07 · Stars 32026-08 · Stars 32026-09 · Stars 326-0326-09

The counting method changed in 2026-09: skills with the same name are now counted separately by author, and the full catalog is covered. Dashed lines show months before the change; they are not compared with later months, and no growth is computed across the change. Other skills share this name; figures from before the change are attributed to this one, which currently has the most downloads.

MonthDownloadsMoMStars
2026-093.9K—3
2026-082.7K+19.6%3
2026-072.2K+45.1%3
2026-061.5K+33.3%1
2026-051.1K—1
2026-03622—1

SKILL.md (original)

The author’s original file, shown as plain text without edits or review by AlphaMoat. Links are not clickable; check commands before running them. Fetched 2026-10-23.

9.0 KB · SkillHub v1.0.0
---
name: wechat-content-studio
description: 微信公众号内容工作室 — 支持多来源权威搜索、多站点文章抓取、AI 改写、封面生成、智能排版发布的一站式工具
author: 模型猎人
version: 2.1.0
allowed-tools: Bash,Read,Write
---

# 微信公众号内容工作室 (WeChat Content Studio)
Show all (215 lines)
一站式公众号内容创作工具,支持**多来源权威搜索**(10 大分类 50+ 来源)和**多站点文章抓取**,自动合并、**AI 改写**、**封面生成**、**智能排版发布**。

默认产出根目录为 **`~/WorkBuddy/<技能文件夹名>/`**(与 `~/.workbuddy/skills` 下本技能目录名一致)。**搜索 / 链接抓取 / workflow**(`search`、`links`、`workflow`)在根目录下再建一层:**合并**时为 `sanitize(合并后标题)/`,**不合并**时为 `1_标题/`、`2_标题/` 子目录,其内为 `article.md`;其它命令(如 `generate-cover`、`extract` 调用多站点抓取)仍按各子命令参数或默认根目录行为。

## 核心工作流

```bash
# 最简用法:多来源搜索 → 抓取 → 合并 → 改写 → 封面 → 排版 → 发布
node {baseDir}/scripts/main.js search "RAG 幻觉治理" --merge
```

## 命令一览

### 1. 多来源搜索(推荐)

```bash
# 默认搜索(高质量渠道:微信全网+知乎+少数派等 + 开发者社区)
node {baseDir}/scripts/main.js search "人工智能" --merge

# 叠加微信公众号头部号定向搜索
node {baseDir}/scripts/main.js search "大模型" --sources high_quality_channels,wechat_top --merge

# 指定搜索分类
node {baseDir}/scripts/main.js search "大模型" --sources wechat_top,intl_ai_official,academic --merge

# 搜索所有 10 大分类
node {baseDir}/scripts/main.js search "AI Agent" --all-sources --merge

# 列出所有可用搜索分类
node {baseDir}/scripts/main.js search "any" --list-sources

# 仅搜索微信公众号头部号(Brave site:mp.weixin.qq.com + 各账号名)
node {baseDir}/scripts/main.js search "RAG" --wechat-only --merge

# 搜索不发布
node {baseDir}/scripts/main.js search "大模型" --merge --no-auto

# 调整每源抓取数和总上限
node {baseDir}/scripts/main.js search "LLM" --count 5 --total-max 30 --merge
```

#### 搜索分类

| 分类 Key | 分类名称 | 包含来源 |
|----------|---------|----------|
| `high_quality_channels` | 高质量内容渠道 | 微信公众号(全网)、知乎、少数派、掘金、InfoQ、极客公园、虎嗅、36氪 |
| `wechat_top` | 微信公众号头部账号 | 机器之心、量子位、AI科技评论、新智元、DataFunTalk、PaperWeekly、深度学习自然语言处理 |
| `cn_tech_blog` | 中文技术博客 | CSDN、博客园、掘金、知乎专栏、SegmentFault、简书、少数派 |
| `cn_bigtech` | 国内大厂官方号 | 阿里达摩院、腾讯AI Lab、百度AI、字节AILab、华为诺亚方舟、智谱AI |
| `intl_ai_official` | 国际头部 AI 公司 | OpenAI Blog、Google AI Blog、Meta AI、Anthropic Blog、Stability AI |
| `intl_media` | 国际权威媒体 | TechCrunch、MIT Technology Review、The Verge、Reuters中文、Bloomberg中文、FT中文、财新 |
| `cn_tech_media` | 国内科技媒体 | 36氪、品玩、极客公园、虎嗅、钛媒体、InfoQ |
| `dev_community` | 开发者社区 | Hacker News (score≥100)、GitHub Trending、Product Hunt、V2EX |
| `platform_blogs` | 平台官方 Blog | GitHub Blog、Hugging Face Blog |
| `academic` | 学术/研究 | arXiv、Papers With Code |
| `investment` | 投资/商业 | 投中网、IT桔子 |
| `cn_tech_other` | 其他中文科技媒体 | 雷锋网、亿欧、CSDN AI专栏 |

默认分类:`high_quality_channels` + `dev_community`(含微信公众号全网与知乎等;可加 `--sources wechat_top` 定向头部号)

#### 搜索引擎

| 来源类型 | 搜索方式 |
|---------|---------|
| 微信公众号(定向头部号) | Brave:`site:mp.weixin.qq.com "公众号名" 关键词` |
| 微信公众号(全网) | Brave:`site:mp.weixin.qq.com 关键词`(类型 `wechat_global`) |
| 知乎(专栏+问答等) | Brave:`site:zhihu.com 关键词` |
| 普通网站(含 CSDN、博客园、掘金等) | Brave Search:`site:domain 关键词` |
| Hacker News | Algolia HN Search API |
| GitHub Trending | HTML 抓取 + 关键词过滤 |
| Product Hunt | Brave `site:producthunt.com` |
| arXiv | OpenSearch API |

### 2. 链接抓取

```bash
node {baseDir}/scripts/main.js links "URL" --merge
node {baseDir}/scripts/main.js links "URL1,URL2" --merge
node {baseDir}/scripts/main.js links --file urls.txt --merge
node {baseDir}/scripts/main.js links "URL" --merge --no-auto
```

### 3. 多站点抓取

```bash
node {baseDir}/scripts/main.js extract "URL"
node {baseDir}/scripts/main.js extract "URL" --publish --theme github
node {baseDir}/scripts/main.js extract "URL" --json
```

支持站点:微信公众号、CSDN、博客园、掘金、知乎、简书、思否、少数派,及通用网页。

### 4. AI 改写

```bash
node {baseDir}/scripts/main.js rewrite ./path/to/article.md
```

### 5. 封面生成

```bash
node {baseDir}/scripts/main.js generate-cover --title "文章标题" --content "内容摘要"
node {baseDir}/scripts/main.js generate-tech --description "RAG 系统架构"
```

### 6. 发布

```bash
node {baseDir}/scripts/main.js publish ./path/to/article.md
node {baseDir}/scripts/main.js publish-browser ./path/to/article.md
```

### 7. 一键工作流

```bash
node {baseDir}/scripts/main.js workflow search \
  --keyword "人工智能" --count 5 --merge --rewrite --generate-cover --publish

node {baseDir}/scripts/main.js workflow links \
  --urls "URL1,URL2" --merge --rewrite --generate-cover --publish
```

## 输出结构

**搜索 / 链接抓取 / workflow**(`--merge`):默认 `--output` 为 `~/WorkBuddy/<技能文件夹名>/`,其下再建 **`文章标题/`**(文件名经 `sanitize`),例如:

```
~/WorkBuddy/wechat-content-studio/
└── <文章标题>/
    ├── article.md              # 合并后的正文(含 frontmatter)
    ├── article_rewritten.md    # AI 改写版
    ├── merged_articles.json    # 合并的原始数据
    ├── images/
    │   └── cover.jpg           # 封面图(若生成)
    └── metadata.json           # 若某流程写出
```

**`--no-merge`**:在同一根目录下建 **`1_标题/`、`2_标题/`** 各含 `article.md`,避免多篇混在同一目录。

## 模块说明

| 模块 | 目录 | 功能 |
|------|------|------|
| 多来源搜索 | `search/multi_source_search.js` | Brave Search(site:)、HN/GitHub/arXiv API |
| 搜索来源配置 | `search/search_sources.json` | 多分类来源(含中文技术博客 CSDN/博客园等) |
| 微信搜索 | — | 已并入 `multi_source_search`(Brave) |
| 文章合并 | `search/merge_articles.js` | 多篇文章智能合并 |
| 抓取 | `extractor/` | 多站点文章提取(通过 multi-site-extractor) |
| 改写 | `rewriter/` | AI 改写、去 AI 味 |
| 配图 | `image/` | 万象 2.6 封面生成(含复用逻辑) |
| 发布 | `publisher/` | wenyan-cli / browser-use 发布 |
| 排版 | 外部依赖 | `wechat-typeset-pro` 排版技能 |

## 环境变量

启动时自动从 `~/.openclaw/.env` 加载,无需手动 export。

| 用途 | 变量名 | 说明 |
|------|--------|------|
| 微信 AppID | `WECHAT_APP_ID` | 发布必须 |
| 微信 AppSecret | `WECHAT_APP_SECRET` | 发布必须 |
| 阿里云 DashScope | `DASHSCOPE_API_KEY` | 封面生成必须 |
| LLM API Key | `OPENAI_API_KEY` | AI 改写必须 |
| LLM Base URL | `OPENAI_BASE_URL` | 默认 `https://api.openai.com/v1` |
| LLM 模型名 | `OPENAI_MODEL` | 默认 `gpt-4o` |
| 联网搜索代理 | `HTTPS_PROXY` / `HTTP_PROXY` | 访问 Brave/GitHub 等;未设时默认 `http://127.0.0.1:7890` |
| Brave 请求间隔 | `BRAVE_SEARCH_MIN_INTERVAL_MS` | 两次 Brave 请求最小间隔(毫秒),默认 `3200`,遇 429 可调大 |

**.env 查找顺序:**
1. `OPENCLAW_ENV_FILE`(若设置)
2. 技能根目录 `.env`
3. `~/.openclaw/.env`
4. `~/.workbuddy/.env`

## 依赖安装

```bash
# Node.js 依赖
cd {baseDir} && npm install

# wenyan-cli(发布功能)
npm install -g @wenyan-md/cli

# Python 依赖(多站点抓取)
pip install requests beautifulsoup4 lxml markdownify readability-lxml

# 浏览器自动化(可选)
uvx browser-use install
```

## 注意事项

1. **发布凭证**:首次使用需配置微信公众号 API 凭证
2. **IP 白名单**:发布功能需将服务器 IP 添加到微信公众号后台白名单
3. **封面复用**:封面生成失败时自动复用上次已上传的封面图
4. **Brave 限流**:短时间大量搜索可能返回 HTTP 429,脚本会退避重试;请降低 `--count`、少用 `--all-sources`,或隔一段时间再搜
5. **版权合规**:生成的文章请确保符合版权法规,建议深度改写

## 技能依赖

- `multi-site-extractor` — 多站点文章提取
- `wechat-typeset-pro` — 专业排版(多主题)
- 阿里云万象 2.6 — 封面图生成
- wenyan-cli / browser-use — 发布到公众号

Popular in Content

Data month: 2026-09 · Downloads, stars and installs are aggregated monthly from public skill registries (ClawHub, SkillHub). See methodology.