Agent skill

WeChat Article Downloader

by Moore-developers in Moore-developers/moore-wechat-article-downloader

Downloads and manages WeChat official account articles from URLs, an account's history or subscriptions, adding engagement metrics and top comments when a valid session exists.

MITAuto-check passedProductivity & Automation

SKILL.md written in Chinese; this summary is our English description.

Install WeChat Article Downloader

skills CLI
$ npx skills add Moore-developers/moore-wechat-article-downloader --skill moore-wechat-article-downloader -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install Moore-developers/moore-wechat-article-downloader moore-wechat-article-downloader --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
moore-wechat-article-downloader
GitHub stars
294
Token cost
~3.4k tokens
SKILL.md length
577 words
Files
40 (incl. scripts, references)
Skills in repo
1
Repo updated
First seen
Licence
MIT

At a glance

Downloads and manages WeChat official account articles from URLs, an account's history or subscriptions, adding engagement metrics and top comments when a valid session exists.

  • Works in 4 steps: 运行 exporter-login-qr-start… → 如果二维码没有自动打开,确认 qrcode_path 文件存在且为 .png /… → 用户确认扫码后运行… → …
  • Downloading a WeChat official account article from its URL
  • SKILL.md covers 场景判断, 范围, 执行护栏 and 场景 1:直接下载已知文章, plus 6 more sections
  • Runs Python scripts from its folder; calls python3; reaches mp.weixin.qq.com

What it does

Written in Chinese, this skill archives articles from WeChat official accounts and prefers to run locally. A wizard script picks the scenario: direct download of full article URLs or a txt, csv or json list; history for a named account through an Exporter login, which is the default, or a proxy fallback; WeChat favorites for comments, reads and likes; and scheduled sync of several subscribed accounts.

The wizard returns gates such as need_login, need_proxy_confirm and need_account_choice, which the agent handles by guiding a QR-code scan, confirming the system proxy or listing candidates in chat. Exporter mode fetches only account searches, history lists and article text, so engagement data comes from a separate batch flow that works only while the user's short WeChat session is valid. After a run the agent checks index.csv, the Markdown, the image folder and page status. It does not bypass logins or paywalls, and does not turn into rewriting, summarizing or a cloud service.

When your agent uses it

  • Downloading a WeChat official account article from its URL
  • Pulling an account's history list and saving the articles locally
  • Saving comments and engagement data for chosen articles
  • Setting up scheduled incremental sync of several accounts

Example prompts

  • “Download all articles from the official account named Tech Weekly and show me the list first.”
  • “Sync the latest five posts from this account and keep the top comments.”
  • “Set up a daily incremental sync for my three WeChat accounts.”

Requirements

  • Python 3 to run the wizard scripts
  • A QR-code login for history sync
  • A valid short WeChat session for engagement data

Workflow steps

4 steps, taken from the first numbered list in SKILL.md.

  1. 运行 exporter-login-qr-start --open;命令会先复用未完成且未过期的二维码会话,只有过期后才生成新二维码,不要省略 --open。
  2. 如果二维码没有自动打开,确认 qrcode_path 文件存在且为 .png / .jpg / .jpeg,再按平台调用系统打开命令。
  3. 用户确认扫码后运行 exporter-login-qr-status,不要靠等待时间猜测状态。
  4. 状态确认后运行 exporter-login-qr-complete,只向用户报告登录是否成功和有效期,不展示 auth-key 或其预览。

What it can do on your machine

Read from SKILL.md and the folder at commit dea919b. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Ships 1 file in scripts/ (Python, from the files we listed), which the agent can run.

    Shell commands in SKILL.md call:

    • python3

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    Hosts in commands or code, which the agent is likely to contact:

    • mp.weixin.qq.com

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

WeChat Article Downloader loads about 3.4k tokens when it runs, and up to ~12k if it reads all its reference files. Until then it costs about 32 tokens; SKILL.md has 577 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~32
When it runs · the whole SKILL.md, loaded when a task matches
~3.4k
With references · SKILL.md plus every file in references/, read only if the agent opens them
~12k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.

SKILL.md

The full file from Moore-developers/moore-wechat-article-downloader at commit dea919b, republished under its MIT licence (© Moore-developers). 577 words, ~3,354 tokens.

Download SKILL.mdSave it as .claude/skills/moore-wechat-article-downloader/SKILL.md (or your agent's skills folder). This skill also uses 39 other files; get the full folder from GitHub.
name
moore-wechat-article-downloader
description
当用户要下载或管理微信公众号文章时使用本技能。支持订阅同步、微信收藏和链接导入;互动数据仅在用户微信短时会话有效时补充指标与精选评论。优先本地运行;不要扩展成内容改写、总结、SaaS 或云服务。

公众号归档器

场景判断

用户意图场景入口
提供了完整文章 URL场景 1:直接下载wechat_wizard.py run "下载:<url>"
给公众号名 / 样例 URL,想要历史文章场景 2A:Exporter(默认)wechat_wizard.py run "历史文章:<名称或URL>"
明确说"不登录"或"用代理"场景 2B:代理(备选)wechat_wizard.py run "代理历史:<url>"
想抓评论/阅读/点赞/页面风格场景 2C:微信收藏wechat_downloader.py proxy-enhancer-session-start --yes
管理多个公众号 / 定时同步场景 3:订阅同步见场景 3

统一入口优先。wizard 返回以下 gate 时按表处理,不要立刻降级为底层命令:

Gate处理
need_login引导扫码 → login-status 轮询 → login-complete 恢复任务
need_proxy_confirm向用户确认启用系统代理 → 确认后继续
need_account_choice聊天中列出候选公众号 → resume --choice "<序号>"
need_article_choice聊天中列出文章标题+日期 → resume --choice "<选择>"

只有 wizard 返回 blocked 或 failed_recoverable 并明确给出底层命令时,才使用底层命令。

范围

可以做:下载公开的 mp.weixin.qq.com 文章、抓取历史文章列表、Exporter 扫码登录和账号管理、定时增量同步。

不要做:绕过登录/付费墙/私密内容;打印 auth-key/cookie/token;把历史列表抓取和 URL 下载混为一谈;扩展成内容改写、总结、SaaS。

执行护栏

  • 不要把命令成功等同于交付完整:success_count 只表示流程没有抛错。结束前必须检查 index.csv、Markdown、图片目录和源页面状态。
  • 不要只给二维码文件路径:生成或复用二维码后优先调用当前操作系统的默认看图程序直接打开。只有无桌面环境或系统打开失败时,才退回聊天图片展示;裸路径只能作为最后兜底。同一未完成登录会话只打开一次,二维码过期后才重新生成。
  • 扫码流程不能少步骤:用户说“已扫码”后先运行 exporter-login-qr-status;只有状态为 confirmed / ready_to_complete=true 才运行 exporter-login-qr-complete。
  • 搜索结果必须精确匹配:公众号搜索可能返回多个相似名称。优先选择昵称完全一致的结果,并核对简介或 alias;不要默认使用第一个模糊匹配结果。
  • 已有明确选择时不要重复询问:用户已经说“最新 N 篇”或给出标题/日期时,该表达本身就是文章选择;同步后直接执行,但结束时仍要在聊天中列出日期、标题和结果。

场景 1:直接下载已知文章

触发:用户提供一篇或多篇完整文章 URL,或 .txt/.csv/.json 文件。

bash
python3 {baseDir}/scripts/wechat_wizard.py run "下载:<url 或文件路径>"

重跑会跳过已成功的 URL。失败项重试:

bash
python3 {baseDir}/scripts/wechat_wizard.py retry "<task-id>"

下载完成后报告:成功数量、失败数量、失败 URL(如有)、output_dir、index.csv。

场景 2:公众号历史文章

2A. Exporter 模式(默认首选)

触发:用户提供公众号名、关键词或样例文章 URL,想要历史文章列表。

优势:API 稳定、无需代理、支持按名称搜索、一次登录 4 天有效、SQLite 持久化可复用。

边界:Exporter 模式只做公众号搜索、历史文章同步和文章下载;不获取评论、阅读数、点赞数、收藏数、转发数等互动数据。互动数据只在代理快照模式验证通过后再提供。

带评论/互动数据的硬规则:

  • 普通下载:exporter-sync -> exporter-download,只用于不带评论/互动数据的正文下载。
  • 带评论互动下载:exporter-sync -> engagement batch download。
  • 当用户需求包含“评论 / 互动 / 阅读数 / 点赞 / 在看 / 精选评论”时,禁止先调用普通 exporter-download。
  • 此时 Exporter 只负责同步文章列表、标题、日期和 URL;正文与互动数据由 wechat-collection-sync-engagement / wizard 的 engagement 批量任务处理。
  • 文章落盘只发生一次:成功拿到互动数据后生成最终 Markdown;如果某篇互动失败,再降级写正文并标记“互动数据缺失”。
bash
python3 {baseDir}/scripts/wechat_wizard.py run "获取公众号「<名称>」的历史文章"

wizard 自动处理:检查登录 → 搜索公众号 → 同步文章 → 列出结果。

文章列表展示(强制要求):

抓取完成后必须在聊天中直接罗列,格式:

text
- **YYYY-MM-DD**:文章标题

不要只给 CSV/JSON 路径。不要只给编号,必须同时展示标题和日期。

让用户按标题关键词、日期、最新 N 篇或编号范围选择后再下载。

2B. 代理模式(备选)

触发:用户明确说"不登录"或"用代理",或 Exporter 模式失败。

bash
python3 {baseDir}/scripts/wechat_wizard.py run "代理历史:<sample-article-url>"

硬性规则:

  • 必须使用旧版历史入口:https://mp.weixin.qq.com/mp/profile_ext?action=home&__biz=<biz>&scene=124#wechat_redirect
  • 不要引导用户使用 channels.weixin.qq.com/web/pages/mp_profile(视频号壳页,不触发 getmsg 接口)
  • 启用系统代理前必须向用户确认;结束时必须用 history-capture-finish 恢复系统代理
  • 默认使用常驻本地代理 127.0.0.1:23344;切换公众号时不要停止代理服务,直接运行新的 history-capture-prepare 复用进程

告诉用户:把旧版入口发到微信文件传输助手,用桌面客户端内置浏览器打开,看到历史列表后向下滚动。

2C. 微信收藏(互动数据与页面保存)

触发:用户要评论、阅读数、点赞数、收藏数、在看数、完整页面快照、公众号排版风格。

此模式保存用户当前浏览的文章,也可在短时凭证有效期内恢复同公众号已授权任务的互动同步。历史列表仍默认走 Exporter;旧代理历史列表只作为备用。只同步精选评论,不承诺全量评论或完整回复树。

批量互动任务必须使用已经同步出的文章列表和 URL 创建,不要先用普通 Exporter 下载正文。用户说“下载某公众号最新 N 篇文章的评论和互动数据”时,正确链路是:

text
exporter-sync -> engagement batch download

主流程不再使用 proxy-snapshot-prepare --yes。

每次新会话从 23032-24045 选择一个空闲随机端口。启动前先固定当前 upstream,再启动增强代理并切换系统 HTTP/HTTPS 代理。端口、PID、upstream 和恢复状态写入活动会话;同一会话安全重载时必须复用这些值。

启动成功后会终止旧的微信 WeChatAppEx WebView 进程。微信主进程和登录状态保持不变;用户下次打开文章时,微信会创建使用当前代理的新 WebView。代理增强代码安全重载后也执行同样的 WebView 重置,避免旧进程继续使用缓存页面。

text
系统/微信 -> 127.0.0.1:<会话端口> -> 会话启动时确定的上游代理或直连 -> 外网

启动命令:

bash
python3 {baseDir}/scripts/wechat_downloader.py proxy-enhancer-session-start --upstream-proxy none --yes

如果修改了代理增强代码或需要重载,只能使用安全重载。该命令自动读取并复用活动端口与 upstream:

bash
python3 {baseDir}/scripts/wechat_downloader.py proxy-enhancer-restart --upstream-proxy none --yes

不要在系统代理指向活动端口时直接 kill、proxy-enhancer-stop 或 stop && start;这会让系统代理指向死端口。安全重载会临时恢复上游或直连,使用原端口重启,再切回活动端口。

启动后常见链路:

text
系统/微信 -> 127.0.0.1:<会话端口> -> 启动时确定的上游代理 -> 外网

以下命令自动读取活动端口,不要手工猜端口:

bash
python3 {baseDir}/scripts/wechat_downloader.py proxy-enhancer-route-help
python3 {baseDir}/scripts/wechat_downloader.py proxy-enhancer-check-ingress --minutes 10
python3 {baseDir}/scripts/wechat_downloader.py proxy-enhancer-logs --hours 24 --limit 80
python3 {baseDir}/scripts/wechat_downloader.py proxy-enhancer-video-links --hours 24 --limit 100

增强代理常驻期间会额外捕获公开的 channels.weixin.qq.com 视频号页面地址,写入 proxy-snapshots/video-links.jsonl。只提供页面链接,不下载视频;禁止保存或展示 finder.video.qq.com 临时媒体地址以及 token、ticket、exportkey、session、auth 等签名凭证。

视频号链接捕获规则:

  • 必须先启动再浏览:常驻拦截只能捕获启动后的流量,不能追溯之前打开过的视频号页面。
  • 先确认流量进入代理:启动后运行 proxy-enhancer-check-ingress;确认微信流量到达,再让用户在微信里打开视频号卡片或页面。
  • 只保存可复用页面链接:仅保留脱敏后的 channels.weixin.qq.com 页面 URL;finder.video.qq.com 通常是短时签名媒体地址,不作为结果。
  • 没有结果不代表没有视频:可能是页面未重新打开、流量未经过代理、链接只存在于客户端内部状态,或公开链接依赖已过滤的临时凭证。报告时必须说明是哪一层没有观察到。
  • 结束时恢复系统代理:用户明确结束常驻拦截后运行 proxy-enhancer-session-finish --yes,避免系统继续指向已停止的本地端口。

视频下载入口:

  • Exporter 直接下载:先运行常驻增强代理并打开目标视频页完成短时媒体凭证捕获,再执行 python3 {baseDir}/scripts/wechat_exporter.py exporter-download --account-id <id> --latest <N> --include-video --video-quality highest。输出在账号目录 videos/ 下,文件名前缀为 [视频]。
  • 视频号页面按钮:常驻增强代理会尝试在 channels.weixin.qq.com/web/pages/... 页面注入“下载视频”按钮;按钮只在用户当前有权限播放、页面 JS 暴露 objectDesc.media[] 后可下载。
  • 无凭证必须明示:Exporter 未发现活跃代理或匹配的短时媒体描述时返回 needs_capture,不得报告视频下载成功。处理方式是保持/启动 proxy-enhancer-session-start --yes,重新打开目标视频页,等按钮显示“下载视频”后再重跑。
  • 敏感字段只留内存:urlToken、decodeKey、签名媒体 URL、Cookie 和请求头只在代理进程内短时保存和下载使用;禁止写入 SQLite、Markdown、JSONL、调试日志或聊天输出。结果只报告页面 URL、descriptor id、文件路径、大小、是否解密。
  • 能力边界:不绕过私密、付费或 DRM 内容;只下载用户当前会话能播放的公开视频。加密视频会按 decodeKey 生成 128 KiB key stream 并 XOR 文件头,失败时标记 failed / decrypt_failed。

调试日志位于 ~/.moore/wechat-article-downloader/proxy-snapshots/debug.jsonl,只记录脱敏事件:请求路径、响应状态、是否识别为文章页、是否注入脚本、页面脚本是否执行、按钮是否放置成功。日志只保留 24 小时;清理动作每 12 小时最多执行一次。

成功识别并注入的文章 HTML 响应必须写入 Cache-Control: no-store,并移除 ETag / Last-Modified,防止微信 WebView 在下一次会话继续复用未更新的注入代码。已禁用的系统代理状态视为没有 endpoint;不要保存或展示其残留的 server/port。

不要在用户点击保存后自动恢复系统代理。只有用户明确要求恢复时才运行:

bash
python3 {baseDir}/scripts/wechat_downloader.py proxy-enhancer-session-finish --yes

该命令先恢复系统代理,再停止并回收本次会话的随机端口进程。

用户之后正常打开公众号文章。页面加载完成后应自动出现注入按钮:顶部元信息行一个,评论区标题旁一个,点击任意一个都会保存当前文章现场。

text
收藏到本地

用户点击后读取最近快照:

bash
python3 {baseDir}/scripts/wechat_downloader.py snapshot-latest
python3 {baseDir}/scripts/wechat_downloader.py snapshot-list --limit 10

如果用户说“已保存”“处理刚才保存的快照”“整理今天保存的快照”,默认不要只处理 latest。先看快照收件箱,再把所有未处理快照归档到文章库:

bash
python3 {baseDir}/scripts/wechat_downloader.py snapshot-inbox --limit 50
python3 {baseDir}/scripts/wechat_downloader.py snapshot-attach --all-unprocessed

归档规则:

  • 已下载过的文章:按 URL/微信文章参数匹配,增强数据放到对应公众号目录。
  • 没下载过的文章:从快照正文生成 Markdown,也放到 ~/Downloads/wechat-articles/<公众号名>/。
  • 评论、阅读数、点赞数、收藏数等页面数据必须写进文章 Markdown 的 ## 页面数据 区块;同一篇文章重复归档时替换旧区块,不重复追加。
  • 同一篇文章多次保存:保留每次快照,并更新 latest.json 和 metrics_history.jsonl。
  • 不递归复制原始快照目录;只归档结构化提取产物,避免把敏感或调试文件带进文章库。

把最近快照提取成后续处理用的结构化文件:

bash
python3 {baseDir}/scripts/wechat_downloader.py snapshot-extract latest

也可以指定快照:

bash
python3 {baseDir}/scripts/wechat_downloader.py snapshot-extract <snapshot-id> --output-dir <dir>

输出位于:

text
~/.moore/wechat-article-downloader/proxy-snapshots/<snapshot-id>/

主要文件:

  • snapshot.json:DOM 快照和基础字段
  • dom.html:脱敏后的完整页面 DOM
  • js_content.html:正文 DOM
  • comments_dom.html:页面已渲染评论 DOM
  • engagement_dom.html:互动区域 DOM
  • metrics.json:阅读、点赞、在看、评论、收藏、分享的可观察结果
  • network.jsonl:脱敏后的网络响应摘要
  • style_profile.json / style_summary.md:页面风格摘要
  • report.md:本次快照报告

snapshot-extract 默认输出到 <snapshot-dir>/extracted/:

  • article.md:正文 Markdown
  • comments.json:已加载评论文本和原始评论 DOM,complete=false
  • comments_structured.json:结构化评论,包含昵称、地区/时间、点赞、回复数、内容
  • metrics.json:可观察互动数据
  • style_profile.json:页面风格特征
  • image_urls.json:正文图片 URL 列表
  • engagement.html:互动区域 DOM
  • report.md:提取报告

规则:不承诺一定拿到所有互动字段;页面未暴露时必须标记 missing。不要保存或打印 cookie、auth-key、token、pass_ticket、key 等敏感值。

文章落盘规则:标题和 Markdown 文件名必须带 [图文] 或 [贴图] 前缀;微信 item_show_type=8 判定为贴图,其他文章判定为图文。贴图 Markdown 先集中输出全部图片,再输出文字内容。

Show full SKILL.md (204 more words)Show less
文章类型与图片完整性
  • 类型判定只看微信字段:item_show_type=8 是贴图;不要根据图片数量、正文长度或是否存在 content_noencode 猜类型。普通图文也可能包含 content_noencode。
  • 正文提取顺序:优先使用非空 #js_content;为空或缺失时安全解码 content_noencode,禁止执行页面 JavaScript。
  • 贴图图片兜底:贴图正文没有 <img> 时,优先读取 share_imageinfo[].cdn_url 中公开的 mmbiz.qpic.cn 原图地址。排除 watermark_info、cover_url、头像、封面和同图不同尺寸派生项,再按规范化 URL 或内容哈希去重。该兜底只用于贴图,避免普通图文误收装饰资源。
  • 标题来源优先级:页面标题有效时使用页面标题;页面返回 untitled、空标题或删除提示时,回退到 Exporter 已同步的数据库标题。frontmatter、index.csv 和文件名必须一致。
  • 删除文章单独标记:页面出现“该内容已被发布者删除”等提示时,不得报告为“全文下载成功”;保留历史标题和 URL,并明确标记正文不可用。
  • 贴图验收:至少检查 content_type=贴图、item_show_type=8、image_count>0、所有本地图片存在,以及第一个图片引用位于正文内容之前。还要抽查图片确为文章内容图,不能只看数量;若 image_count=0,必须继续排查元数据图片清单,不能直接结束。
  • 图文验收:检查标题/文件名 [图文] 前缀、正文非空、Markdown 图片引用对应本地文件。图片下载失败要在 error 中保留原因。

下载后建议按以下顺序验收:

text
命令结果 -> index.csv 类型/数量 -> Markdown 标题与排序 -> 图片文件存在 -> 删除/风控页面检查 -> 聊天中报告真实结果

场景 3:多账号订阅 + 定时同步

触发:用户需要长期跟踪多个公众号,希望每天自动同步并下载新文章。

已有能力:

bash
# 搜索并添加公众号
python3 {baseDir}/scripts/wechat_exporter.py exporter-search "关键词"
python3 {baseDir}/scripts/wechat_exporter.py exporter-add --fakeid "<fakeid>" --nickname "<name>"

# 查看已订阅的公众号列表
python3 {baseDir}/scripts/wechat_exporter.py exporter-accounts

# 同步单个账号文章
python3 {baseDir}/scripts/wechat_exporter.py exporter-sync --account-id "<id>" --limit 200

# 下载指定账号最新 N 篇
python3 {baseDir}/scripts/wechat_exporter.py exporter-download --account-id "<id>" --latest 20
bash
# 同步所有已订阅账号(默认每个账号最新 50 篇)
python3 {baseDir}/scripts/wechat_exporter.py exporter-sync-all
python3 {baseDir}/scripts/wechat_exporter.py exporter-sync-all --per-account-limit 100

# 下载所有尚未下载的新文章
python3 {baseDir}/scripts/wechat_exporter.py exporter-download-new

# 每日一键运行:sync-all + download-new(适合 cron)
python3 {baseDir}/scripts/wechat_exporter.py exporter-daily-run

设置 cron 定时任务时,用 AI 帮用户生成 launchd plist 或 crontab 条目,指定绝对路径调用 exporter-daily-run。

Exporter 登录

QR 流程(首选):

bash
python3 {baseDir}/scripts/wechat_exporter.py exporter-login-qr-start --open
python3 {baseDir}/scripts/wechat_exporter.py exporter-login-qr-status "<login-id>"
python3 {baseDir}/scripts/wechat_exporter.py exporter-login-qr-complete "<login-id>"

降级:手动 auth-key(QR 不可用时):

bash
python3 {baseDir}/scripts/wechat_exporter.py exporter-login-start --open
python3 {baseDir}/scripts/wechat_exporter.py exporter-config --auth-key "<auth-key>"

规则:扫码时必须选公众号/服务号,不要选小程序;auth-key 有效期约 4 天;不在聊天中打印完整 auth-key;auth-key 优先存 macOS Keychain。

二维码交互要求:

  1. 运行 exporter-login-qr-start --open;命令会先复用未完成且未过期的二维码会话,只有过期后才生成新二维码,不要省略 --open。
  2. 如果二维码没有自动打开,确认 qrcode_path 文件存在且为 .png / .jpg / .jpeg,再按平台调用系统打开命令。
  3. 用户确认扫码后运行 exporter-login-qr-status,不要靠等待时间猜测状态。
  4. 状态确认后运行 exporter-login-qr-complete,只向用户报告登录是否成功和有效期,不展示 auth-key 或其预览。

系统打开命令:

环境首选命令备用命令
macOSopen "<qrcode_path>"open "file://<absolute-path>"
Windows PowerShellStart-Process -FilePath "<qrcode_path>"explorer.exe "<qrcode_path>"
Windows CMDstart "" "<qrcode_path>"explorer.exe "<qrcode_path>"
Linux 桌面xdg-open "<qrcode_path>"gio open "<qrcode_path>"
WSLexplorer.exe "$(wslpath -w '<qrcode_path>')"cmd.exe /c start "" "$(wslpath -w '<qrcode_path>')"

执行规则:

  • 本地路径优先:二维码是本地图片,直接打开路径比转换成网页 URL 更稳定。
  • 禁止拼接不可信 shell 字符串:代码实现应使用参数数组调用进程;Windows Python 优先使用 os.startfile(),避免 cmd /c start 的引号陷阱。
  • 无桌面环境要降级:SSH、容器或 Linux 没有 DISPLAY / WAYLAND_DISPLAY 时不要反复调用打开命令,改为聊天展示二维码图片。
  • 打开后仍要提示动作:系统窗口弹出后明确告诉用户“请用微信扫码并在手机端确认”,不能只执行命令后沉默。

本地管理页(仅用户明确要求时启动):

bash
python3 {baseDir}/scripts/wechat_exporter.py exporter-server-start --port 8765

路径和输出

{baseDir} = 本 SKILL.md 所在目录,脚本位于 {baseDir}/scripts/。

内部会话数据:~/.moore/wechat-article-downloader/

默认下载目录:

  • URL 模式:~/Downloads/wechat-articles/<公众号名>/;多公众号 URL 会按公众号拆分目录
  • Exporter 单公众号模式:~/Downloads/wechat-articles/<公众号名>/
  • Exporter 多公众号模式:按公众号拆分为多个 ~/Downloads/wechat-articles/<公众号名>/

输出结构:

text
index.csv
articles/<safe-title>.md
images/<safe-title>/<image-number>.<ext>

URL 模式和 Exporter 模式都不再使用可见的 <run-id> 交付目录。run_id 只保存在运行记录里。Exporter 下载会先检查账号目录里的 index.csv 和实际 Markdown/图片文件;SQLite 标记为已下载但文件缺失时会重新下载。

输出约定

每次结束报告:使用的模式、成功/失败数量、失败 URL(如有)、output_dir、index.csv。

报告中的“成功”必须区分:正文完整、图片完整、源文章已删除。不要把 HTTP 请求成功或 Markdown 文件生成成功描述成“全文和图片均已获取”。

Exporter 模式额外报告:auth-key 是否有效、公众号/文章数量、SQLite 路径。

历史文章模式:必须先在聊天中列出文章标题和日期,让用户选择,不要自动全量下载。

保持回复简洁,详细内容由文件承载。

参考文件

  • references/backend-design.md:两模式架构
  • references/skill-cli-flow.md:Skill 驱动 CLI 流程
  • references/output-formats.md:输出文件结构
  • references/compliance.md:安全和权限规则
  • references/troubleshooting.md:常见失败和处理

© Moore-developers, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 39 other files (scripts, references) in the repository root of Moore-developers/moore-wechat-article-downloader.

  • SKILL.md
  • .gitignore
  • LICENSE
  • README.md
  • docs/plans/2026-07-06-unified-wizard-gates.md
  • docs/plans/2026-07-08-skill-scenario-redesign.md
  • docs/plans/2026-07-09-proxy-snapshot-validation-plan.md
  • docs/plans/2026-07-10-wechat-collection-architecture.md
  • docs/plans/2026-07-10-wechat-collection-implementation-tasks.md
  • evals/evals.json
  • evals/fixtures/exporter_fixture.json
  • evals/fixtures/html/demo-url-article-1.html
  • evals/fixtures/html/demo-url-article-2.html
  • evals/fixtures/html/demo-url-article-content-noencode.html
  • evals/test_credential_broker.py
  • evals/test_exporter_contract.py
  • … and 24 more

Open the folder on GitHubat commit dea919b

Compare with similar skills

WeChat Article Downloader next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

WeChat Article Downloader compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
WeChat Article Downloader this skillMoore-developers/moore-wechat-article-downloader294—~3.4kAutomated safety check: PassMIT
Junta Leiloeirossickn33/agentic-awesome-skills47k2 repos~1.6kAutomated safety check: PassMIT
Skyvern Browser AutomationSkyvern-AI/skyvern23k—~1.9kAutomated safety check: PassAGPL-3.0
Create Experkfly/ex-skill2.5k—~3.7kAutomated safety check: NotesMIT
WeChat Article Searchzjp1997720/wechat-article-search2681 repos~730Automated safety check: NotesMIT
Cron Opsczl9707/build-your-own-openclaw1.9k—~593Automated safety check: PassMIT

Similar skills

  • Junta Leiloeiros

    sickn33/agentic-awesome-skills

    Coleta e consulta dados de leiloeiros oficiais de todas as 27 Juntas Comerciais do Brasil.

    47k GitHub starsUsed in 2 repos~1.6k tokens
    Backend & APIsAuto-check passed
  • Skyvern Browser Automation

    Skyvern-AI/skyvern

    Automates websites with Skyvern's AI browser agent to fill forms, extract data, download files, log in and run multi-step workflows through SDKs, REST, MCP or a CLI.

    23k GitHub stars~1.9k tokensUpdated today
    Productivity & AutomationAuto-check passed
  • Create Ex

    perkfly/ex-skill

    Distill an ex-girlfriend into an AI Skill. An agent skill from perkfly/ex-skill.

    2.5k GitHub stars~3.7k tokensUpdated 6 mo ago
    Productivity & AutomationAuto-check: notes
  • WeChat Article Search

    zjp1997720/wechat-article-search

    Searches WeChat public account articles by keyword through Sogou WeChat Search and returns titles, summaries, dates, source accounts and links.

    268 GitHub starsUsed in 1 repo~730 tokens
    Productivity & AutomationAuto-check: notes
  • Cron Ops

    czl9707/build-your-own-openclaw

    Create, list, and delete scheduled cron jobs. An agent skill from czl9707/build-your-own-openclaw.

    1.9k GitHub stars~593 tokensUpdated 3 mo ago
    Productivity & AutomationAuto-check passed
  • Wx4 Skill

    claw-codes/wx4py

    微信自动化操作 skill,用于帮助用户快速完成微信消息群发、文件发送、群管理、聊天记录获取、多群监听、AI 自动回复等自动化任务。当用户需要批量发送微信消息、管理微信群、获取聊天记录、监听群消息、或接入 AI 群聊机器人时使用此 skill。

    758 GitHub stars~3k tokensUpdated 6 mo ago
    Productivity & AutomationAuto-check passed

Questions about WeChat Article Downloader

What does WeChat Article Downloader do?

Downloads and manages WeChat official account articles from URLs, an account's history or subscriptions, adding engagement metrics and top comments when a valid session exists. Written in Chinese, this skill archives articles from WeChat official accounts and prefers to run locally. A wizard script picks the scenario: direct download of full article URLs or a txt, csv or json list; history for a named account through an Exporter login, which is the default, or a proxy fallback; WeChat favorites for comments, reads and likes; and scheduled sync of several subscribed accounts.

When should I use WeChat Article Downloader?

WeChat Article Downloader fits situations like: downloading a WeChat official account article from its URL; pulling an account's history list and saving the articles locally; saving comments and engagement data for chosen articles; setting up scheduled incremental sync of several accounts.

How do I install WeChat Article Downloader in Claude Code?

Run `npx skills add Moore-developers/moore-wechat-article-downloader --skill moore-wechat-article-downloader -a claude-code`. Or copy the skill folder (the Moore-developers/moore-wechat-article-downloader repository) into .claude/skills/moore-wechat-article-downloader in your project. Claude Code loads it when a task matches its description.

How do I install WeChat Article Downloader in Codex?

Run `npx skills add Moore-developers/moore-wechat-article-downloader --skill moore-wechat-article-downloader -a codex`. Or copy the skill folder (the Moore-developers/moore-wechat-article-downloader repository) into .agents/skills/moore-wechat-article-downloader in your project. Codex loads it when a task matches its description.

Can I use WeChat Article Downloader in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add Moore-developers/moore-wechat-article-downloader --skill moore-wechat-article-downloader -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/moore-wechat-article-downloader, .gemini/skills/moore-wechat-article-downloader, .github/skills/moore-wechat-article-downloader and .opencode/skills/moore-wechat-article-downloader in your project.

What does WeChat Article Downloader need to run?

Going by SKILL.md and its folder, WeChat Article Downloader needs Python for the scripts in its folder and the command-line tools its instructions call (python3). Our summary lists: Python 3 to run the wizard scripts; A QR-code login for history sync; A valid short WeChat session for engagement data.

Does WeChat Article Downloader access the network?

SKILL.md names 1 domain. In commands or code: mp.weixin.qq.com; the agent is likely to contact it when it follows the instructions. This is read from the text; nothing was executed.

Is WeChat Article Downloader safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.

What licence does WeChat Article Downloader use?

WeChat Article Downloader is published under the MIT licence (from the LICENSE file in the skill folder). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does WeChat Article Downloader use?

About 3.4k tokens (SKILL.md is roughly 13k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 8.3k tokens, read only when the agent opens those files.

What are the alternatives to WeChat Article Downloader?

Skills that share tags, products or a category with WeChat Article Downloader: Junta Leiloeiros (sickn33/agentic-awesome-skills, 47k stars), Skyvern Browser Automation (Skyvern-AI/skyvern, 23k stars), Create Ex (perkfly/ex-skill, 2.5k stars) and WeChat Article Search (zjp1997720/wechat-article-search, 268 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains WeChat Article Downloader?

Moore-developers (a GitHub user) maintains it in Moore-developers/moore-wechat-article-downloader, which has 294 GitHub stars. The repository was last updated on August 1, 2026.

Source: Moore-developers/moore-wechat-article-downloader on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.