Linkdigest Social Link Reader
sickn33/agentic-awesome-skills
Read one public Xiaohongshu, Douyin, TikTok, YouTube, X or WeChat article link into text an agent can use (transcript, image text, key points) via the LinkDigest API or MCP server.
使用 VideoDevour 把视频(B站/YouTube/抖音/X 链接、微信视频号分享链接或本地文件)处理成中文图文报告。当用户要求"处理这个视频"、"视频转笔记/报告/图文大纲"、"下载并总结B站/YouTube/抖音/X/视频号视频"时使用。支持搜索视频、查询链接信息、一键生成带关键帧的图文报告(精简/详细),改写成量子速读/公众号文章/小红书笔记、导出…
$ npx skills add datawhalechina/video-devour --skill videodevour -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install datawhalechina/video-devour videodevour --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/datawhalechina/video-devour.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.agents/skills/videodevour .claude/skills/videodevour && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "videodevour" agent skill from https://github.com/datawhalechina/video-devour/tree/main/.agents/skills/videodevour into .claude/skills/videodevour/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "videodevour", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/datawhalechina/video-devour/tree/main/.agents/skills/videodevourType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add datawhalechina/video-devour --skill videodevour -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install datawhalechina/video-devour videodevour --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/datawhalechina/video-devour.git skills-src && mkdir -p .agents/skills && cp -r skills-src/.agents/skills/videodevour .agents/skills/videodevour && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "videodevour" agent skill from https://github.com/datawhalechina/video-devour/tree/main/.agents/skills/videodevour into .agents/skills/videodevour/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "videodevour", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add datawhalechina/video-devour --skill videodevour -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install datawhalechina/video-devour videodevour --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/datawhalechina/video-devour.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/.agents/skills/videodevour .cursor/skills/videodevour && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "videodevour" agent skill from https://github.com/datawhalechina/video-devour/tree/main/.agents/skills/videodevour into .cursor/skills/videodevour/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "videodevour", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/datawhalechina/video-devour.git --path .agents/skills/videodevour--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add datawhalechina/video-devour --skill videodevour -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install datawhalechina/video-devour videodevour --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/datawhalechina/video-devour.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/.agents/skills/videodevour .gemini/skills/videodevour && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "videodevour" agent skill from https://github.com/datawhalechina/video-devour/tree/main/.agents/skills/videodevour into .gemini/skills/videodevour/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "videodevour", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install datawhalechina/video-devour videodevourInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add datawhalechina/video-devour --skill videodevour -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/datawhalechina/video-devour.git skills-src && mkdir -p .github/skills && cp -r skills-src/.agents/skills/videodevour .github/skills/videodevour && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "videodevour" agent skill from https://github.com/datawhalechina/video-devour/tree/main/.agents/skills/videodevour into .github/skills/videodevour/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "videodevour", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add datawhalechina/video-devour --skill videodevour -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install datawhalechina/video-devour videodevour --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/datawhalechina/video-devour.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/.agents/skills/videodevour .opencode/skills/videodevour && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "videodevour" agent skill from https://github.com/datawhalechina/video-devour/tree/main/.agents/skills/videodevour into .opencode/skills/videodevour/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "videodevour", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
videodevour使用 VideoDevour 把视频(B站/YouTube/抖音/X 链接、微信视频号分享链接或本地文件)处理成中文图文报告。当用户要求"处理这个视频"、"视频转笔记/报告/图文大纲"、"下载并总结B站/YouTube/抖音/X/视频号视频"时使用。支持搜索视频、查询链接信息、一键生成带关键帧的图文报告(精简/详细),改写成量子速读/公众号文章/小红书笔记、导出…
Videodevour is an agent skill from datawhalechina/video-devour. 使用 VideoDevour 把视频(B站/YouTube/抖音/X 链接、微信视频号分享链接或本地文件)处理成中文图文报告。当用户要求"处理这个视频"、"视频转笔记/报告/图文大纲"、"下载并总结B站/YouTube/抖音/X/视频号视频"时使用。支持搜索视频、查询链接信息、一键生成带关键帧的图文报告(精简/详细),改写成量子速读/公众号文章/小红书笔记、导出 PDF,生成学习测试题(单选/多选/判断,判题与学习评估),并可检索项目本地已积累的文档库(历史任务报告/笔记,支持 BM25 搜索与导出,也可通过 MCP 服务供其他 LLM 调用)。
Its SKILL.md is about 3k tokens, which your agent loads only when the skill is triggered. The skill folder holds 2 other files, including scripts (for example `scripts/devour.py`). Compatibility notes: 需要 Python 3.12+ 与项目 .venv(uv sync),ffmpeg;任何支持 .agents/skills 约定的 agent 均可调用
It sits in Documents & Office, covering PDF. It works with YouTube, Model Context Protocol, Douyin and X (Twitter). The repository describes itself as: 🚀 基于 ASR + VLM 技术的智能视频笔记工具,能够将任何视频"吞噬"并生成包含图文内容和视频剪影的结构化笔记报告. The licence is Apache-2.0.
12 steps, taken from the step headings in SKILL.md.
Read from SKILL.md and the folder at commit 487eddf. It shows what the files ask for, not the result of running them.
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.
Ships 1 file in scripts/ (Python), which the agent can run.
Shell commands in SKILL.md call:
python3uvbashpythonFrom the folder's file list and the shell code blocks in SKILL.md.
Hosts in commands or code, which the agent is likely to contact:
bilibili.comdouyin.comweixin.qq.comv.douyin.comx.comFrom URLs in SKILL.md, links to its own repository left out.
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
需要 Python 3.12+ 与项目 .venv(uv sync),ffmpeg;任何支持 .agents/skills 约定的 agent 均可调用
From compatibility in the SKILL.md frontmatter.
Videodevour loads about 3k tokens when it runs. Until then it costs about 72 tokens; SKILL.md has 548 words of instructions outside code blocks.
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.
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.
The full file from datawhalechina/video-devour at commit 487eddf, republished under its Apache-2.0 licence (© datawhalechina). 548 words, ~3,037 tokens.
.claude/skills/videodevour/SKILL.md (or your agent's skills folder). This skill also uses 1 other file; get the full folder from GitHub.调用 VideoDevour 项目(本仓库),把视频端到端处理为中文图文报告 (ASR 转写 → 大纲 → 视频切分 → 关键帧 → 图文报告),不依赖 Web 界面。 产出三层内容:图文大纲 / 精简报告(结论先行)/ 详细报告(完整原文+笔记对照); 可按需改写成量子速读 / 公众号文章 / 小红书笔记,并导出 PDF。 项目同时把每次任务的产物沉淀为本地文档库,可跨任务检索复用(见第 11 节)。
--home 参数 → VIDEO_DEVOUR_HOME 环境变量 → 脚本所在仓库 → 默认安装路径。cd <项目目录> && uv sync,用 Python 3.12);
脚本会自动切换到 .venv 运行。settings.json;LLM/VLM 需要 key。
ASR 默认走在线(云端识别,零模型下载);仅当用户要用本地离线 ASR 时,才需下载
约 2GB 模型:设置页切到「离线」会显示自检结果与安装命令,或执行
bash scripts/install_offline_asr.sh。缺 key 时提示用户在 WebUI 控制台(/settings)填写。reportlab(已写入 requirements.txt / requirements-lite.txt),
执行过 uv sync 即可用。以下 <skill目录> 指本 SKILL.md 所在目录(.agents/skills/videodevour)。
# B站(默认)
python3 <skill目录>/scripts/devour.py search "关键词" --platform bilibili --max 5
# YouTube
python3 <skill目录>/scripts/devour.py search "关键词" --platform youtube --max 5
# 抖音(需要登录 Cookie,见第 4 节)
python3 <skill目录>/scripts/devour.py search "关键词" --platform douyin --max 5输出 JSON 列表(标题/时长/UP主/链接),交给用户选择或选最匹配的一条。
a_bogus 浏览器签名校验,若接口触发人机校验,脚本会给出「改用浏览器搜索后
粘贴链接」的引导,而非静默返回空结果。python3 <skill目录>/scripts/devour.py info "https://www.bilibili.com/video/BV..."输出标题/UP主/时长/封面 JSON。注意:多P合集的 BV 链接,info 返回的时长是合集总时长,
而 process 只下载并处理当前分P(通常几分钟),不会被合集时长吓退。
# 首次使用先检查元宝 Cookie(视频号解析依赖腾讯元宝接口登录态)
python3 <skill目录>/scripts/devour.py wechat --check
# 下载视频号视频到项目 uploads/(仅下载,返回 JSON:file_path/title/uploader)
python3 <skill目录>/scripts/devour.py wechat "https://weixin.qq.com/sph/..."
# 也可直接粘贴含链接的分享文案;下载后接 process 本地文件即得图文报告--check 会明确返回并给出配置指引;配置方法见 README
「微信视频号 Cookie 配置」一节(WebUI 设置页或 settings.json / 环境变量均可)process 命令同样直接支持视频号链接(自动先下载再处理),wechat 用于只要视频文件的场景# 检查抖音登录 Cookie 是否已配置
python3 <skill目录>/scripts/devour.py douyin --check
# 下载抖音视频到项目 uploads/(仅下载,返回 JSON:file_path/title/uploader)
python3 <skill目录>/scripts/devour.py douyin "https://www.douyin.com/video/7624464836100967732"
# 短链与分享文案均可(自动跟随重定向、提取链接)
python3 <skill目录>/scripts/devour.py douyin "https://v.douyin.com/xxxxxx/"settings.json 的 douyin_cookies。
Cookie 缺失或失效时,脚本返回明确的配置指引。aweme.snssdk.com):web 端接口自 2026 起受 Argus
浏览器签名校验拦截,仅带登录 Cookie 也会返回 403 Blocked by ArgusSecurityPlugin,
yt-dlp 会把它误报成「Fresh cookies are needed」。看到该提示时不要反复重配 Cookie,
先确认下载是否已走 App 接口成功;仅当 App 接口也失败时才需要重新读取登录态。douyin.com/video/{id} 视频页、douyin.com/note/{id} 图文、
含 ?modal_id={id} 的搜索页链接、v.douyin.com 短链(脚本自动归一化)。download_addr;
需要干净画面(如关键帧配图)时设 VIDEO_DEVOUR_DOUYIN_CLEAN=1,改为优先无水印源
(分辨率可能更低)。process 命令同样直接支持抖音链接(自动先下载再处理);douyin 用于只要视频文件的场景。# 下载 X 推文视频到项目 uploads/(仅下载)
python3 <skill目录>/scripts/devour.py process "https://x.com/{user}/status/{id}"
# 或只下载不处理:info 查看 / process 处理,均支持 X 链接auth_token);也可写入
settings.json 的 x_cookies。未登录时 yt-dlp 常返回「No video could be found」。x.com / twitter.com、/{user}/status/{id}、/i/status/{id}、/i/web/status/{id}、
mobile.twitter.com、带 ?s= 追踪参数等形态(脚本自动归一化)。process 命令同样直接支持 X 链接(自动先下载再处理)。不下载视频、不走 ASR:直接读取平台已有字幕(B站 AI 字幕轨 / YouTube 手动或自动字幕), LLM 整理为纯文本要点笔记。适合"只要文字内容、要快"的场景。
python3 <skill目录>/scripts/devour.py notes "https://www.bilibili.com/video/BV..." --level 高中输出 JSON:notes 为 Markdown 笔记全文(含 ```mermaid 概念关系图)、
file(.txt 纯文本)/ md_file(.md 含关系图)落盘路径。注意:视频没有字幕轨时
报错并建议改走 process 完整流程;YouTube 无 cookies 或被 bot 检查拦截时提示配置
「YouTube cookies」(设置控制台,可一键读取浏览器 Cookie)。
python3 <skill目录>/scripts/devour.py process "https://www.bilibili.com/video/BV..." --level 高中
# 或抖音 / 视频号 / 本地文件:
python3 <skill目录>/scripts/devour.py process "https://www.douyin.com/video/..." --level 高中
python3 <skill目录>/scripts/devour.py process /path/to/video.mp4 --level 初中<数据根>/downloads/,映射表
index.json 记录来源 URL/平台/标题/命中次数),不重复下载;换链接形式(短链/搜索页链接)
也能命中同一缓存(按「平台+视频ID」为键)。report:精简报告(结论先行,快速阅读)detailed_report:详细报告(完整视频原文 + 整理笔记对照,原文不截断)outline:图文大纲(章节 + 关键帧)keyframes:关键帧图片路径数组media_profile:压缩前后体积与分辨率(确认存储收益时看这个)styles:已生成的衍生文体(未生成时不出现,见第 8 节)settings.json(默认在线 DashScope/StepFun;离线=本地 MPS/CUDA/CPU,
需先装模型),LLM/VLM 固定走云端VIDEO_DEVOUR_FRAMES_PER_CHAPTER
可调),再由 VLM 评分挑选每章最佳一张——不随视频长短暴增候选帧--extras "mindmap,graph,card"(可任选其一或多个,逗号分隔)在报告完成后一次生成;
也可事后用 mindmap / graph 子命令单独生成。不要在用户未要求时主动生成基于任务报告生成两种知识可视化(LLM 生成、浏览器渲染的交互页面):
# 思维导图:三层分支结构(markmap 渲染,可缩放/折叠)
python3 <skill目录>/scripts/devour.py mindmap --latest --open
# 知识图谱:概念关系力导向网络(节点按类别着色、关系标注)
python3 <skill目录>/scripts/devour.py graph --latest --open
# 也可用 --dir 指定任务输出目录、--level 指定学习阶段输出 JSON {"html": "<输出目录>/mindmap.html", ...},将 html 路径呈现给用户(浏览器打开即用)。
若在 process 时已用 --extras 生成过,直接使用输出 JSON 中 extras 里的路径,无需重复生成。
把已有报告改写成可直接发布的成品文案(不是分析报告,基于精简报告改写,事实不跑偏)。 不在处理流程里预生成,按需生成并落盘缓存,重复调用直接复用。
S=<skill目录>/scripts/devour.py
# 三种全部生成(默认)
python3 $S styles --latest
# 只生成某一种
python3 $S styles --latest --kind xiaohongshu
# 指定任务目录
python3 $S styles --dir "<输出目录>" --kind quantum,wechatquantum 量子速读:30 秒看懂大意 + 一句可直接发朋友圈的话wechat 公众号文章:图文成稿(保留关键帧配图),可直接发布xiaohongshu 小红书笔记:emoji 分点的图文笔记,含话题标签输出 JSON {"styles": {"quantum": {"label": "...", "file": "..."}, ...}}。
用户没要求就不要主动生成(每次约 10-30 秒)。
python3 $S pdf --latest # 精简报告(默认)
python3 $S pdf --latest --type detailed # 详细报告
python3 $S pdf --latest --type all # 大纲+精简+详细 合并
python3 $S pdf --latest --out ~/Desktop/report.pdf # 指定输出路径服务端用 reportlab 排版(内置中文字体,不依赖浏览器/pandoc),关键帧图片内嵌。
输出 JSON {"pdf": "...", "type": "...", "bytes": N}。同样按需使用。
处理完成后读取打印的 report / detailed_report 路径,向用户呈现内容摘要。
默认看精简报告;要看完整原文对照用 --type detailed:
python3 <skill目录>/scripts/devour.py report --latest # 精简报告
python3 <skill目录>/scripts/devour.py report --latest --type detailed # 详细报告(原文+笔记)项目把每次任务的产物沉淀为本地文档库(含 WebUI 历史 + skill 直跑的任务)。 当用户想要的内容之前已处理过、或问题可以在已有笔记里找到答案时,先检索文档库, 再决定是否重新处理视频,能省下大量时间。
S=<skill目录>/scripts/devour.py
# BM25 相关度检索(中英文混合),返回 doc_id/scope/相关度/摘要
python3 $S library search "沙箱隔离" --scope all --top 5
# 无关键词浏览全部(可按 scope 过滤)
python3 $S library list --scope report
# 按 doc_id + scope 取全文 Markdown(直接打印)
python3 $S library get <doc_id> --scope report
# 同一视频多次处理有多个版本时用 --run-id 指定
python3 $S library get <doc_id> --scope detailed --run-id <run_id>
# 导出单篇为 ZIP(md + 引用图片) / 导出整库
python3 $S library export <doc_id> --scope report --out ~/Desktop
python3 $S library export-all --out ~/Desktopscope 取值:all(默认)/ outline 图文大纲 / report 精简报告 / detailed 详细报告 /
quantum 量子速读 / wechat 公众号文章 / xiaohongshu 小红书笔记doc_id、run_id、scope、label、version_label(同一视频多版本)、
source_url(原视频链接)、score、snippet;取全文时把 doc_id + scope(多版本再加
run_id)传给 library get 即可。如果用户希望 Claude Code / Claude Desktop / Cursor 等 MCP 客户端直接检索本地文档库,
无需通过本 skill 脚本:项目内置 stdio MCP 服务 mcp_server/videodevour_library_mcp.py,
暴露 5 个工具:search_library / get_article / list_library / export_article / export_library。
接入配置(写入客户端的 mcpServers):
{
"mcpServers": {
"videodevour-library": {
"command": "<项目目录>/.venv/bin/python",
"args": ["<项目目录>/mcp_server/videodevour_library_mcp.py"]
}
}
}command 用项目 .venv 的 python,args 指向服务脚本绝对路径python mcp_server/test_mcp_client.py(逐个调用全部工具并校验导出副作用)依据任务报告生成测试习题(单选题 / 多选题 / 判断题)检验学习掌握情况:
服务端判题(答案不下发,防偷看)、逐题解析、分章掌握度、历次成绩,并可按需生成
LLM 学习建议(基于本次错题)。试卷落盘 quiz.json 复用,作答记录落盘 quiz_attempts.json。
S=<skill目录>/scripts/devour.py
# 出题(默认 10 题;--count 调整,--force 重新出题)。输出不含答案的题目列表
python3 $S quiz generate --latest --count 10
# 向用户呈现题目收集作答后判题:answers 格式 {"题目id": [选项下标...]}(多选多个下标)
python3 $S quiz grade --answers '{"q1":[0],"q2":[0,2]}' --latest
# 逐题对错/正确答案/解析 + 总分 + 分章掌握度;grade 输出含 attempt_id
# 基于本次错题生成学习建议(LLM,只生成一次并缓存)
python3 $S quiz advice --attempt-id <grade输出的attempt_id> --latest默认(最快路径,适合只要报告的用户):
S=<skill目录>/scripts/devour.py
python3 $S process "https://www.bilibili.com/video/BV..." --level 高中 # 下载+处理,产出报告
python3 $S report --latest # 精简报告全文先查库、避免重复处理(用户的问题可能已有现成笔记):
S=<skill目录>/scripts/devour.py
python3 $S library search "用户的关键词" --top 5 # 命中 → library get 取全文
# 没有命中 → 再走 process 处理视频要发内容时(在已有报告基础上改写,按需生成):
S=<skill目录>/scripts/devour.py
python3 $S styles --latest --kind quantum # 量子速读(速览 + 朋友圈文案)
python3 $S styles --latest --kind wechat # 公众号图文
python3 $S pdf --latest --type detailed # 导出 PDF全套学习材料(仅当用户明确需要导图/图谱/卡片时):
S=<skill目录>/scripts/devour.py
python3 $S process "https://www.bilibili.com/video/BV..." --level 高中 --extras "mindmap,graph,card"检验学习效果(用户想自测时,先出题再判题):
S=<skill目录>/scripts/devour.py
python3 $S quiz generate --latest # 出题 → 呈现给用户作答
python3 $S quiz grade --answers '...' --latest # 判题 + 评估,必要时再 quiz advice向用户交付时建议按「报告全文 → 思维导图 → 知识图谱」的顺序呈现:先细节后框架,便于学习理解。
<项目>/output/frames_*/ 目录,与 WebUI 历史共用(skill 直跑的任务不注册到 WebUI 任务列表)./start.sh 后访问
http://localhost:8000;默认监听 0.0.0.0,同一局域网的其他电脑用 http://<本机IP>:8000 也能打开WECHAT_RESOLVER_URL);处理前可用 wechat --check 验证 Cookie;
失败时引导用户用本地捕获工具(ltaoo/wx_channels_download)下载后按本地文件处理。
视频号不支持搜索,只能粘贴分享链接a_bogus 浏览器签名校验,触发时改用「浏览器搜索 → 复制链接 → 粘贴处理」。下载固定选
H.264 格式(H.265 直链会 403)© datawhalechina, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
SKILL.md and 1 other file (scripts) in .agents/skills/videodevour of datawhalechina/video-devour.
Open the folder on GitHubat commit 487eddf
Videodevour 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.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Videodevour this skilldatawhalechina/video-devour | 157 | — | ~3k | Automated safety check: Pass | Apache-2.0 | |
| Linkdigest Social Link Readersickn33/agentic-awesome-skills | 47k | — | ~3.4k | Automated safety check: Pass | MIT | |
| PullmdAeternaLabsHQ/pullmd | 486 | — | ~2.6k | Automated safety check: Pass | AGPL-3.0 | |
| Lecture To Notesysyecust/lecture-to-notes | 273 | — | ~14k | Automated safety check: Notes | Custom licence | |
| Read URLs and PDFstw93/Waza | 7.2k | — | ~1.8k | Automated safety check: Pass | MIT | |
| Web To Markdownrookie-ricardo/erduo-skills | 935 | — | ~894 | Automated safety check: Pass | MIT |
sickn33/agentic-awesome-skills
Read one public Xiaohongshu, Douyin, TikTok, YouTube, X or WeChat article link into text an agent can use (transcript, image text, key points) via the LinkDigest API or MCP server.
AeternaLabsHQ/pullmd
Read any web page, document, or YouTube video as clean Markdown using PullMD.
ysyecust/lecture-to-notes
A skill your agent uses when users provide YouTube, Bilibili, or X/Twitter lecture URLs and want reader-first Chinese LaTeX/PDF notes with source-faithful claims, fluent authored prose, and verified…
tw93/Waza
Fetches web pages and PDFs and returns a source-grounded summary, clean Markdown, quotes or citations, routing each kind of link to a suitable fetch method.
rookie-ricardo/erduo-skills
Convert a web URL into cleaned Markdown with deterministic routing.
SerhiiKorniienko/bullshit-detector
Fetch and normalize any content source into clean text with metadata — YouTube video transcripts, TikTok captions, web articles, PDFs, tweets/X posts, local files.
datawhalechina/video-devour
VideoDevour 桌面客户端(macOS .app / Windows 安装包)的打包、签名、发版与热更新。当用户要求"打包客户端"、"构建/出安装包"、"发版/发布 release"、"实现热更新或自动更新"、"客户端装不上/打开报错/签名问题"时使用;也用于排查打包特有的故障——应用被判"已损坏"、功能在开发模式正常但客户端里失效、模块缺失(No module named…
Categories
使用 VideoDevour 把视频(B站/YouTube/抖音/X 链接、微信视频号分享链接或本地文件)处理成中文图文报告。当用户要求"处理这个视频"、"视频转笔记/报告/图文大纲"、"下载并总结B站/YouTube/抖音/X/视频号视频"时使用。支持搜索视频、查询链接信息、一键生成带关键帧的图文报告(精简/详细),改写成量子速读/公众号文章/小红书笔记、导出…. Videodevour is an agent skill from datawhalechina/video-devour.
Videodevour fits situations like: tasks that involve PDF.
Run `npx skills add datawhalechina/video-devour --skill videodevour -a claude-code`. Or copy the skill folder (.agents/skills/videodevour in datawhalechina/video-devour) into .claude/skills/videodevour in your project. Claude Code loads it when a task matches its description.
Run `npx skills add datawhalechina/video-devour --skill videodevour -a codex`. Or copy the skill folder (.agents/skills/videodevour in datawhalechina/video-devour) into .agents/skills/videodevour in your project. Codex loads it when a task matches its description.
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add datawhalechina/video-devour --skill videodevour -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/videodevour, .gemini/skills/videodevour, .github/skills/videodevour and .opencode/skills/videodevour in your project.
Going by SKILL.md and its folder, Videodevour needs Python for the scripts in its folder and the command-line tools its instructions call (python3, uv, bash and python). Our summary lists: Python 3. Compatibility (from SKILL.md): 需要 Python 3.12+ 与项目 .venv(uv sync),ffmpeg;任何支持 .agents/skills 约定的 agent 均可调用.
SKILL.md names 5 domains. In commands or code: bilibili.com, douyin.com, weixin.qq.com, v.douyin.com and x.com; the agent is likely to contact these when it follows the instructions. This is read from the text; nothing was executed.
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.
Videodevour is published under the Apache-2.0 licence (declared in SKILL.md). It allows redistribution, so the full SKILL.md is shown on this page.
About 3k tokens (SKILL.md is roughly 12k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.
Skills that share tags, products or a category with Videodevour: Linkdigest Social Link Reader (sickn33/agentic-awesome-skills, 47k stars), Pullmd (AeternaLabsHQ/pullmd, 486 stars), Lecture To Notes (ysyecust/lecture-to-notes, 273 stars) and Read URLs and PDFs (tw93/Waza, 7.2k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
datawhalechina (a GitHub organization) maintains it in datawhalechina/video-devour, which has 157 GitHub stars. The repository holds 2 skills in this directory. The repository was last updated on September 21, 2026.
Source: datawhalechina/video-devour on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.