Wechat Oa

微信公众号草稿箱管理工具集。触发词（满足任一即触发）：看看草稿箱/查看草稿/草稿列表/公众号草稿/搜草稿/搜索草稿/按关键词找草稿/按标题搜/创建草稿/新建草稿/发文章到公众号/推送文章/更新草稿/删除草稿/批量删除草稿/生成封面图/上传图片到公众号/上传图片到素材库/已发布文章列表/公众号素材列表/素材管理/删...

Security Scan

The skill's code and runtime instructions match its WeChat drafting purpose, but metadata and SKILL.md contain inconsistent claims about dependencies, credentials, and homepage — you should verify the source and necessary local configuration before installing.

Purpose & Capability

The code (wechat_push.py + generate_infographic.py) implements WeChat draft/material management and image generation which fits the stated purpose. However the SKILL.md and README claim "no third‑party dependencies" while the code clearly uses external Python packages (requests, Pillow, optional premailer). The registry metadata lists no required env vars or primary credential, but the tool requires AppID/AppSecret stored in config.json. These mismatches between claims and actual needs are concerning.

Instruction Scope

Runtime instructions require copying/editing config.json (containing AppID/AppSecret), adding your server IP to WeChat's IP whitelist, and reading design.md before generating HTML. The instructions also suggest querying an external endpoint (curl ifconfig.me) to find your public IP. Other than that, the code's network calls appear limited to official WeChat API endpoints. The instructions do require reading local files (design.md, config.json) and writing generated images to disk, which is expected for this tool.

Install Mechanism

No install spec is provided (instruction-only in registry), so the skill will not install automatically. However the README instructs installing Python packages (requests, Pillow) and the code imports premailer optionally — these dependencies are not declared in the registry metadata. Because there is no automated install step, users must manually install packages; ensure you install them from trusted package indexes and in an isolated environment.

Credentials

The tool legitimately requires WeChat credentials (AppID and AppSecret) to call official APIs, and it stores them in a local config.json. But the skill metadata does not declare these credentials or mark a primary credential, which is inconsistent and can lead to accidental credential exposure. The number and nature of credentials requested (AppID/AppSecret) are proportional to the purpose, but the omission from the declared requirements is a red flag.

Persistence & Privilege

The skill does not request always:true and is user-invocable only. It reads/writes local files (config.json, generated images) and does network calls to WeChat APIs; that level of persistence and privilege is appropriate for a client-side WeChat management tool.

What to consider before installing

This skill appears to implement what it says (WeChat draft/material management and image generation) but contains metadata/instruction mismatches you should address before use: - Verify the source: the registry 'Source' is unknown while SKILL.md/README reference a GitHub repo. Confirm the code origin and trustworthiness before running. - Credentials: The tool requires your WeChat AppID and AppSecret stored in config.json. Do NOT commit that file to any remote repo. Only provide these credentials in a secure environment and consider using least-privilege accounts. - Dependencies: Manually install required Python packages (requests, Pillow; premailer is optional) in a virtual environment. The registry did not declare these — the skill will fail otherwise. - IP whitelist: If your WeChat account uses IP whitelisting, you'll need to add the machine's public IP (instructions reference curl ifconfig.me). Be aware this queries an external service to get your public IP. - Safety: Review the two included Python files yourself (or have a trusted reviewer do so). Network calls in the code target official WeChat endpoints; there are no obvious hidden remote endpoints, but the metadata inconsistencies lower trust. If you plan to use this skill in an autonomous agent, remember the agent will be able to run the code and thus access any local config.json containing credentials — only enable autonomous invocation if you trust the skill and its origin.

微信公众号草稿箱管理工具集。基于官方微信 API，无需第三方依赖。

WeChat Official Account draft management toolkit. Built on official WeChat APIs, no third-party dependencies required.

特性 / Features

• 公众号排版规范 / WeChat MP layout specification：内置 design.md 排版规范，AI 生成 HTML 时必须遵循，确保公众号渲染兼容 / Built-in design.md layout spec that AI must follow when generating HTML, ensuring WeChat rendering compatibility
• 行内样式转换 / Inline-style conversion：自动将 HTML 中的 <style> 标签转换为行内 style="" 属性，兼容微信文章渲染（已固化到 skill） / Auto-converts <style> tags into inline style="" attributes for WeChat-compatible rendering (baked into this skill)
• 自动封面生成 / Auto cover generation：根据文章标题 AI 生成科技风封面图（2.35:1 比例） / AI-generated tech-style cover image from article title (2.35:1 ratio)
• 正文图片自动上传 / Auto inline image upload：自动提取 HTML/MD 中的本地图片，上传到微信素材库并替换 URL / Automatically extracts local images from HTML/MD, uploads to WeChat material library and replaces URLs
• 智能摘要 / Smart digest：自动从正文中选取含功能关键词的段落作为摘要，而非机械截取前80字 / Intelligently selects keyword-rich paragraphs as digest instead of blindly truncating at 80 chars

⚠️ 排版规范（必读） / Layout Specification (MUST READ)

创建或更新公众号文章前，AI 必须先阅读 design.md，严格按其规范生成 HTML。

Before creating or updating any WeChat article, AI MUST read design.md and strictly follow its rules to generate HTML.

design.md 规范涵盖 / design.md covers:

• 容器宽度（文章 677px / 图文卡片 375px）
• 字体规范（clamp() 响应式字号）
• 配色规范（≤5 主体色，对比度 ≥4.5:1）
• 布局规范（Flex/Grid only，禁止 absolute/float）
• 标题规范（禁止重复主标题、header 标签、左边框装饰）
• 内容结构（扁平结构，有/无背景色的 padding 规则不同）
• CSS/HTML/JS 限制（公众号渲染器兼容性白名单）

支持的创作类型 / Supported creation types:

• 文章：通用类型，页面默认宽度 677px
• 贴图：图文卡片（小绿书/小红书风格），页面默认宽度 375px，固定分页比例（默认 3:4）

功能与对应API / Commands & APIs

命令 Command	说明 Description	底层API Underlying API
list	查看草稿列表（含标题+时间）/ View draft list with title+time	draft/batchget
find <关键词>	按标题关键词搜索草稿 / Search drafts by title keyword	draft/batchget
create <文件>	创建新草稿（支持 .html 和 .md，自动上传正文配图）/ Create draft from HTML or MD (auto-upload inline images)	draft/add + 永久素材上传
update <media_id> <文件>	更新已有草稿（自动上传正文配图）/ Update existing draft (auto-upload inline images)	draft/update
update <media_id> <文件> --force-cover	更新草稿并强制重新生成封面 / Update draft + force-regenerate cover	draft/update
delete <media_id>	删除草稿 / Delete draft	draft/delete
batch-del <id1> [id2] ...	批量删除草稿 / Batch delete drafts	draft/delete
upload <图片文件>	上传图片到永久素材库 / Upload image to permanent material	material/add_material
materialcount	获取各类永久素材总数 / Get permanent material count by type	material/get_materialcount
materials [type] [count] [offset] [keyword]	批量获取素材列表，支持关键词过滤 / Batch list with keyword filter	material/batchget_material
materialdel [media_id...]	批量删除素材（多个空格分隔）/ Batch delete materials	material/del_material
materialdel	交互式删除（列出→选编号→确认）/ Interactive material deletion	material/del_material
materialdel <type>	交互式删除指定类型 / Interactive deletion by type (news/image/video/voice)	material/del_material
published	获取已发布文章列表 / List published articles	material/batchget_material
cover <标题>	生成封面图预览（不推送）/ Generate cover preview (no push)	PIL local generation
infographic <类型> <输出路径> [参数]	生成正文配图（流程图/对比图/时间线/文字卡片/统计图）/ Generate inline image (steps/comparison/timeline/textcard/stats)	PIL local generation
userinfo <openid>	获取用户基本信息（需认证账号）/ Get user info by openid	user/info
userlist [next_openid]	获取用户列表（需认证账号）/ List all users	user/get

初始化配置 / Initial Setup

使用前必须完成以下配置：Complete the following before first use:

1. 获取 AppID 和 AppSecret / Get AppID & AppSecret

(1) 登录 微信公众平台 / Log in to WeChat Official Account Platform (2) 进入 设置与开发 → 基本设置 / Go to Settings & Development → Basic Settings (3) 复制 公众号 AppID 和 公众号 AppSecret（如未设置需先启用）/ Copy AppID and AppSecret (enable if not set yet)

2. 添加 IP 白名单 / Add IP Whitelist

调用微信 API 前，必须将服务器 IP 加入白名单：You must add your server's public IP to the whitelist before calling WeChat APIs:

(1) 进入 设置与开发 → 安全中心 / Go to Settings & Development → Security Center (2) 点击 IP 白名单 / Click IP Whitelist (3) 添加你的服务器公网 IP（查看当前 IP：curl ifconfig.me）/ Add your server's public IP (check: curl ifconfig.me)

⚠️ 不添加 IP 白名单会导致 API 调用失败！ / Not adding the IP whitelist will cause all API calls to fail!

3. 配置凭证 / Configure Credentials

将 config.example.json 复制为 config.json，填入你的凭证：Copy config.example.json to config.json and fill in your credentials:

cp config.example.json config.json
# 然后编辑 config.json，填入 APP_ID 和 APP_SECRET
# Then edit config.json, fill in APP_ID and APP_SECRET

config.json 示例 / Example:

{
  "APP_ID": "wx0000000000000000",
  "APP_SECRET": "00000000000000000000000000000000",
  "author": "龙虾"  // 文章默认作者 / Default article author
}

⚠️ config.json 包含凭证，不要提交到 GitHub！已在 .gitignore 中忽略。 / config.json contains credentials — do NOT commit to GitHub! Already in .gitignore.

使用示例 / Usage Examples

# 查看草稿列表 / View draft list
python wechat_push.py list

# 创建新草稿（自动生成封面）/ Create new draft (auto-generate cover)
python wechat_push.py create article.html

# 更新已有草稿 / Update existing draft
python wechat_push.py update n2BZd2CzoCKkl... article.html

# 更新草稿 + 强制重绘封面 / Update draft + force regenerate cover
python wechat_push.py update n2BZd2CzoCKkl... article.html --force-cover

# 删除草稿 / Delete draft
python wechat_push.py delete n2BZd2CzoCKkl...

# 上传图片到永久素材 / Upload image to permanent material
python wechat_push.py upload cover.png

# 生成封面图预览（不推送到微信）/ Generate cover preview (no WeChat push)
python wechat_push.py cover "你的文章标题"

# 生成正文配图 / Generate inline infographic
python generate_infographic.py steps output/step.png "步骤1" "步骤2" "步骤3"
python generate_infographic.py comparison output/compare.png "优点:很好用" "缺点:有点贵"
python generate_infographic.py timeline output/timeline.png "2024:事件1" "2025:事件2"
python generate_infographic.py textcard output/quote.png "天下没有难汇的款"
python generate_infographic.py stats output/stats.png "满意度:85" "便利性:90"

依赖 / Dependencies

pip install requests Pillow

正文配图说明 / Inline Images Guide

自动上传流程 / Auto-upload Flow

创建或更新草稿时，系统会自动处理正文中的图片：

When creating or updating drafts, the system automatically processes inline images:

HTML/MD 文件
    ↓
提取 <img src="..."> 或 ![alt](path)
    ↓
本地图片？ ──是──→ 上传到微信素材库 ──→ 获取微信 URL
    ↓ 否                    ↓
网络图片？ ──是──→ 保留原 URL（可选下载上传）
    ↓ 否
跳过（已是微信素材库图片）
    ↓
替换 HTML 中的 src 为微信 URL
    ↓
推送草稿

支持的图片格式 / Supported Formats

• HTML: <img src="local/image.png"> 或 <img src="http://example.com/img.png">
• Markdown: ![描述](./images/photo.jpg)
• 路径类型: 相对路径、绝对路径、file:/// 协议

图片处理规则 / Image Processing Rules

图片类型	处理方式	说明
本地图片（相对/绝对路径）	自动上传	上传后替换为微信素材库 URL
网络图片（http/https）	跳过	保留原 URL（微信可能过滤）
微信素材库图片	跳过	已是 mmbiz.qpic.cn 域名
不存在的图片	报错提示	记录到失败列表，继续处理其他图片

使用示例 / Examples

<!-- HTML 示例：本地图片会被自动上传 -->
<p>请看下图：</p>
<img src="./images/diagram.png" alt="架构图">
<img src="C:\Users\Photos\screenshot.jpg">

<!-- 网络图片保留原样（微信可能过滤） -->
<img src="https://example.com/external.png">

<!-- Markdown 示例 -->
![本地图片](./assets/chart.png)  ← 自动上传
![网络图片](https://site.com/img.jpg)  ← 保留原样

注意事项 / Notes

1. 图片大小: 建议单张 < 2MB，微信素材库有容量限制
2. 图片格式: 支持 JPG、PNG、GIF，推荐使用 PNG
3. 路径问题: 相对路径基于 HTML/MD 文件所在目录解析
4. 失败处理: 上传失败的图片会记录但不会影响草稿创建

正文配图自动生成 / Auto Infographic Generation

功能说明

generate_infographic.py 可以根据章节内容自动生成配图，无需 AI API，完全本地 PIL 生成。

支持的配图类型

类型	用途	示例
steps	流程图	汇款步骤：注册→填表→汇款→完成
comparison	对比图	传统汇款 vs 西联汇款
timeline	时间线	2020→2022→2024 发展历程
textcard	文字卡片	金句、要点提炼
stats	数据统计图	各渠道手续费对比

使用方式

# 流程图（步骤流程）
python generate_infographic.py steps output/step.png "注册西联账号" "填写汇款信息" "完成汇款" "通知收款人"

# 对比图（两列对比）
python generate_infographic.py comparison output/compare.png "传统银行:3-5个工作日" "西联:几分钟到账"

# 时间线
python generate_infographic.py timeline output/timeline.png "2020:全球扩张" "2022:数字化转型" "2024:移动端上线"

# 文字卡片（金句/要点）
python generate_infographic.py textcard output/quote.png "天下没有难汇的款"

# 数据统计
python generate_infographic.py stats output/stats.png "手续费:最低" "到账速度:最快" "覆盖范围:最广"

配图插入文章流程

1. 生成配图: 运行 generate_infographic.py 生成图片
2. 插入 HTML: 在章节末尾添加 <img src="配图路径">
3. 推送: 运行 python wechat_push.py create article.html
4. 自动上传: 系统自动将本地图片上传到素材库

<h2>三、汇款流程</h2>
<p>以下是完整的汇款步骤：</p>
<img src="./images/remit-steps.png" alt="汇款流程图">  <!-- 自动上传 -->

配图尺寸

• 宽度: 677px（与公众号内容宽度一致）
• 高度: 根据内容自动计算
• 格式: PNG（透明背景支持）
• 字体: Windows 微软雅黑 / 黑体

输出文件 / Output Files

• draft_ids.txt - 草稿记录（创建时间、标题、media_id）/ Draft log (timestamp, title, media_id)
• 封面图默认保存在 HTML 文件同目录 / Cover images saved in the same directory as the HTML file by default

问题反馈 / Feedback

遇到问题或功能建议？欢迎通过以下方式联系我们：

• 🌐 GitHub Issues：https://github.com/andy8663/wechat-oa
  • 提交 Bug 报告或功能请求 / Submit bug reports or feature requests
• 📧 邮箱：[e-mail]
  • 邮件咨询 / Email for inquiries
• 🔔 微信公众号：技术定义未来（ID: gh_b906288c4c2f）
  • 公众号文章首发平台 / Primary publishing platform for MP articles

💡 提交 Issue 前建议先搜索是否已有类似问题 / Please search existing issues before submitting new ones.