REA
Reverse engineer anything from your terminal or agent with one CLI and MCP server.
面向 AI Agent 与终端的全能下载与归档工具 / All-in-one CLI & Skill to download anything: videos, docs, cloud drives, music, papers & books
If you made this, paste this code into your README. The badge links back to this page.
[](https://mcp.popnx.com/en/skills/nengong-download-anything)MCP servers that connect the AI to services for the same work
Reverse engineer anything from your terminal or agent with one CLI and MCP server.
Codebase intelligence for TypeScript and JavaScript, exposed as structured MCP analysis tools.
MCP server for terminal commands, file operations, and process management
Same kind of work, ordered like the Skills list
| This one download-anything | rea | andrej-karpathy-skills | graphify | MoneyPrinterTurbo | |
|---|---|---|---|---|---|
| Maker | nengong | morluto | multica-ai | Graphify-Labs | harry0703 |
| Skills | 1 | 1 | 1 | 1 | 1 |
| GitHub stars | 258 | 468,733 | 234,332 | 150,287 | 144,566 |
| Last code change | Nov 8, 2026 | Nov 19, 2026 | Apr 20, 2026 | Nov 18, 2026 | Nov 18, 2026 |
| License | – | MIT | – | Apache-2.0 | MIT |
1 skills (instruction sets) by nengong that an AI reads and follows · not an MCP server
MCP = the hands
Connects the AI to an outside service or data, such as reading Notion or creating an invoice. A program or server has to be running.
Skill = the manual
A text file that teaches the AI how to do a task. It connects to nothing.
Some overlap: a few Skills run scripts that call tools, and a few MCP servers only send knowledge. Start from the task you want done.
Claude CodeCodex
From the maker’s README and the folders in the repo
GitHub account nengong (person) is only 4 months old: take extra care
★ 258 stars, few users so far · code last changed Nov 8, 2026
1 scripts that can run on a computer: scripts/install.py
Scripts that go online: scripts/install.py
No hooks or writes to machine settings found
Checked automatically from the repo files at commit f3db928 (read only, nothing was run). Not a safety guarantee: read it and decide for yourself
First: which AI do you use?
Pick your app and this page shows only how to install it there
This skill has no license, so we do not hand out a ZIP from our site. Download it from GitHub and arrange the folder yourself, or use "Let the AI install it" above, which is easier
Download from GitHub (whole repo) ↗
Unzip it, then zip only the skills/download-anything folder again. The ZIP must have a download-anything folder with SKILL.md at the top level, no loose files
+Create skillUpload a skillpick download-anything.zip
SettingsCapabilitiesCode execution and file creation
Turn on Code execution and file creation first (newer menus: Create and edit files) · personal plans: Settings → Capabilities · Team/Enterprise: an admin turns it on under Organization settings → Plugins & skills · every plan, including Free
Upload once; it works in claude.ai and Claude Desktop, and in recent Claude Code signed in with a Claude account (not with an API key)
Cowork and cloud sessions use the skills turned on in your claude.ai account, not ~/.claude/skills on your computer (Claude Desktop scheduled tasks run locally and still load it)
Checked against official docs Nov 12, 2026 · Source: support.claude.com, support.claude.com, code.claude.com
Copy this sentence and paste it into a Claude Code chat
For one project, paste it into a chat that is open in that project’s folder · The AI always shows you the path before installing
The AI tells you what it will do first and asks your permission at every step
Restart Claude Code, then type /download-anything or ask for a matching task and Claude picks it up
Checked against official docs Nov 12, 2026 · Source: code.claude.com
A Skill is instructions the AI will follow. Read the files before installing, and install only from makers you trust.
Already installed? Check first · Open Terminal, type this command and press Enter
node -vMac: press ⌘ + Space, type Terminal, press Enter
Windows: open Start, type Windows PowerShell, then Open
Not installed: Open the official download page, pick the LTS version and the click-through installer: .pkg on Mac · .msi on Windows. Double-click and follow it to the end. No curl command needed
Open the Node.js download page ↗
When it’s done, close Terminal, open it again and run the same command. You should see a version number: node -v
Install once; it works for everything that needs Node.js
Checking the commands · copy buttons appear once they pass
Copied from the maker’s README; it downloads and runs the maker’s code
Mac: press Command+Space and type Terminal · Windows: the command below does not work in PowerShell; use “Let AI install it” above instead
This downloads the repo to a temporary folder and copies only the folders that contain a SKILL.md
Checking the commands · copy buttons appear once they pass
The ChatGPT desktop app has Codex built in, and Codex reads skills from ~/.agents/skills
Open Codex in the ChatGPT desktop app and paste this sentence into the chat
The AI always shows you the path before installing
The AI tells you what it will do first and asks your permission at every step
Put the download-anything folder at ~/.agents/skills/download-anything
~ is your home folder ($HOME)
A Skill is instructions the AI will follow. Read the files before installing, and install only from makers you trust.
Mac: press Command+Space and type Terminal · Windows: the command below does not work in PowerShell; use “Let AI install it” above instead
This downloads the repo to a temporary folder and copies only the folders that contain a SKILL.md
Checking the commands · copy buttons appear once they pass
Checked against official docs Nov 15, 2026 · Source: learn.chatgpt.com
Skills you install yourself do not work here; only skills that come with a plugin from the directory
To use this skill with ChatGPT, use the ChatGPT desktop app as above
Checked against official docs Nov 15, 2026 · Source: learn.chatgpt.com
Copy this sentence and paste it into a Codex chat · Works in Codex inside the ChatGPT desktop app, Codex CLI and the IDE extension
For one project, paste it into a chat that is open in that project’s folder · The AI always shows you the path before installing
The AI tells you what it will do first and asks your permission at every step
Codex picks up newly installed skills by itself; if one does not show up, restart Codex · To use it: in ChatGPT type @ and pick the skill · in Codex CLI or the IDE extension run /skills or type $ and the skill name
Checked against official docs Nov 15, 2026 · Source: learn.chatgpt.com
A Skill is instructions the AI will follow. Read the files before installing, and install only from makers you trust.
Mac: press Command+Space and type Terminal · Windows: the command below does not work in PowerShell; use “Let AI install it” above instead
This downloads the repo to a temporary folder and copies only the folders that contain a SKILL.md
Checking the commands · copy buttons appear once they pass
Copy this sentence and paste it into a Cursor chat
For one project, paste it into a chat that is open in that project’s folder · The AI always shows you the path before installing
The AI tells you what it will do first and asks your permission at every step
In Cursor’s Agent chat, type / and pick download-anything
Checked against official docs Nov 12, 2026 · Source: cursor.com
A Skill is instructions the AI will follow. Read the files before installing, and install only from makers you trust.
Mac: press Command+Space and type Terminal · Windows: the command below does not work in PowerShell; use “Let AI install it” above instead
This downloads the repo to a temporary folder and copies only the folders that contain a SKILL.md
Checking the commands · copy buttons appear once they pass
For other apps that read skills. The sentence gives the paths for Claude Code, Codex and Cursor so the AI picks the one for your app · Copy this sentence into the app’s chat
For one project, paste it into a chat that is open in that project’s folder · The AI always shows you the path before installing
The AI tells you what it will do first and asks your permission at every step
Checked against official docs Nov 12, 2026 · Source: code.claude.com, learn.chatgpt.com, cursor.com
Other apps that support Agent Skills (such as Gemini CLI or GitHub Copilot) use different folders. Check that app’s guide, then copy the folder with the SKILL.md there.
Skill folders checked against the official Claude Code, Codex and Cursor guides on Nov 11, 2026
The text below comes from the maker’s repo, not from PopNx MCP
Showing the first 8,000 characters · Open the full file on GitHub
---
name: download-anything
description: 面向个人与 AI Agent 的全能下载与结构化归档工具。当用户要求下载、提取、转换或归档社交视频与图文(B站及分P合集、抖音、小红书、YouTube、X、微博、Instagram、TikTok、m3u8)、播客与音频(小宇宙、Apple Podcasts、网易云音乐、Spotify、Apple Music、本地 NCM 解密与 FLAC 转 MP3)、文章与知识库(微信公众号、知乎/掘金/CSDN/Medium 等技术社区、Obsidian 同步)、学术论文与公开课课件(arXiv、DOI 论文、MIT OCW 开放讲义 PDF、ISBN/MD5 电子书)、开源商用字体(霞鹜文楷、思源黑体/宋体、阿里普惠体、JetBrains Mono、Inter 等)、百度/夸克网盘分享、GitHub 单文件与发行包或微软商店/安卓 APK 安装包时触发;也可在用户要求总结视频、公开课、播客或文章链接时调用,以提取本地字幕(.srt)、Shownotes 或 Markdown 正文供 Agent 分析。
---
# Download Anything (DA)
调用 `da`(或 `dl`)将分享口令文本、链接、DOI、ISBN、MD5、包名或本地音频文件下载并结构化归档到本地。全量核心下载、解密、格式转换与完整性校验均 **开箱即用、免配置本地运行**。各平台详细输入输出规范见 [references/platforms.md](references/platforms.md),开源字体实测梯队与直链见 [references/fonts.md](references/fonts.md),公开课归档与 `not_for` 边界见 [references/education.md](references/education.md),完整 CLI 参数与 JSON 字段定义见 [references/cli.md](references/cli.md)。
## 1. 支持的能力与产物速查(按平台合并,单入口自动识别)
- **视频、图文与社交**(B站 / 抖音 / 小红书 / YouTube / X / 微博 / Instagram / TikTok / `.m3u8`):自动识别视频(输出 `.mp4`,抖音支持免水印最高 4K;加 `-a` 提取纯音轨 `.m4a`/`.mp3`)或多图(`image_01.jpg` + `companion_files`);自动生成 `cover.jpg` 原画封面、`content.md`、`content.info.json`;**B站支持 `?p=N` 分 P 精准选集与 `.srt` 字幕提取**,**抖音提取 `audio.mp3` 原声**,**小红书实况图提取配套 `.mov`**。
- **播客、音乐与音频**(小宇宙 / Apple Podcasts / 网易云音乐 / Spotify / Apple Music / 本地 `.ncm` 与 `.flac`):输出 `.m4a` 或 `.mp3`(本地转换统一输出 `320kbps .mp3`);配套生成播客 `content.md`(完整 Shownotes 与时间轴)、音乐 `.lrc` 歌词、高清专辑封面。
- **文章与知识库**(微信公众号 / 知乎 / 掘金 / CSDN / 博客园 / Medium / Substack / 少数派 / 简书 / SegmentFault / 开源中国 / V2EX):输出干净 Markdown `article.md`,防盗链配图全部本地化(`img_01.png`...)并重写相对路径;加 `--vault` 自动注入 YAML frontmatter 同步至 Obsidian。
- **学术论文、公开课课件与电子书**(arXiv 链接或编号 / DOI / MIT OCW `ocw.mit.edu` 课程讲义与 `.edu` 开放课件 PDF / 10或13位 ISBN / 32位 MD5):输出经 `pypdf` 页数校验的论文/讲义 `.pdf` 或电子书 `.epub` / `.pdf`;论文与公开课讲义自动生成标准 `citation.bib` 引用文件与包含作者/摘要的 `content.md`(商业付费/DRM 网课边界见 [references/education.md](references/education.md))。
- **网盘分享**(百度网盘 / 夸克网盘):自动识别内嵌提取码,支持子目录递归查找与云端临时目录自动清理;ZIP/TAR 默认安全解压并记录于 `extracted_files`。
- **代码、开源字体与安装包**(GitHub `blob` 单文件与 `releases` 发行包 / 开源商用字体 `.ttf`/`.otf`/`.woff2`/`.zip` / 微软商店链接或 12 位 Product ID / Android APK 包名或链接):输出原代码文件、经 `font-magic` 魔数校验的字体文件(ZIP 字体包自动安全解压,实测梯队见 [references/fonts.md](references/fonts.md))、`.msixbundle`(限免费 MSIX/Appx 应用)或 `.apk`(基于 F-Droid/Aptoide 镜像),配套 `content.md` 提供安装说明。
## 2. AI Agent 执行工作流(重要)
### 第一步:环境检查与安装确认(首次调用必检)
在执行下载前,**先确认本机是否已安装 `da` 及其运行依赖,未安装时必须先询问用户确认,不要静默修改系统环境**:
1. **定位可执行文件**:优先查找 `PATH` 中的 `da`(或 `dl`),若无则检查默认独立虚拟环境 `~/.download-anything/venv/bin/da`(Windows 为 `%USERPROFILE%\.download-anything\venv\Scripts\da.exe`)。
2. **未安装 `da` 或缺少 Python 3.10+ 时(先询问用户)**:
- 若上述路径均不存在,先检查本机是否有 `python3`(需 `3.10` 及以上版本;注意 macOS 系统自带 `python3` 常为 `3.9`)。
- **主动向用户询问确认**:告知用户本机尚未初始化 `da` 运行环境(或缺少 Python 3.10+),询问是否允许运行本 Skill 目录下的安装脚本 `python3 <Skill目录>/scripts/install.py`(若无 Python 3.10+,脚本可通过 `uv` 下载独立的 Python 3.12 并创建 `~/.download-anything/venv` 隔离环境,不污染系统 Python)。**待用户同意后**,再执行 `python3 <Skill目录>/scripts/install.py --yes`。
3. **运行环境自检(`da doctor --json`)**:
- 首次安装后或遇到环境报错时,运行 `da doctor --json` 检查 `all_required_ok` 与 `media_ready`。
- **缺少 Python 依赖包**:若 `all_required_ok` 为 `false`,询问用户后运行 `da doctor --fix --json` 自动补齐(官方 PyPI 超时会自动切换清华/阿里云镜像)。
- **音视频任务缺少 `ffmpeg` / `ffprobe`**:若当前任务涉及视频或音频,而 `media_ready` 为 `false`(未安装 FFmpeg),请明确告知用户“文档、图片、论文、电子书与网盘功能不受影响,但音视频合并、转码与双轨校验需要 FFmpeg”,并**询问用户是否需要代为执行安装命令**(macOS:`brew install ffmpeg`;Windows:`winget install -e --id Gyan.FFmpeg`;Linux:`sudo apt install ffmpeg`)。
4. **无头环境(Headless)登录态配置**:
- 在远程服务器、CI/CD 或 Docker 容器等无桌面/无浏览器环境中,交互式 `da login <平台>` 无法调用本地桌面浏览器或系统钥匙串。
- **环境变量注入(推荐容器/CI)**:通过安全方式配置 `DA_<PLATFORM>_COOKIE` 或 `<PLATFORM>_COOKIE`(如 `DA_BAIDU_COOKIE`、`QUARK_COOKIE`)。Instagram 多图使用 `DA_INSTAGRAM_COOKIE_FILE` 指向授权 Cookie 文件,不使用上述 Cookie 字符串变量。
- **挂载或预置 credentials.yaml**:在用户目录配置 `~/.download-anything/credentials.yaml`(权限设为 `0600`,如 `chmod 600 ~/.download-anything/credentials.yaml`),按平台填入 Cookie。
- **安全红线**:严禁在 Agent 的终端输出或聊天对话中打印明文登录凭据、会话 Cookie 或密码。
### 第二步:前置转换与工具边界判断
当用户只提供名称而非直接链接/ID 时,**Agent 应先自行检索出精确标识符,再传给 `da`**:
- **电子书**:若用户只给书名(如“帮我下《傲慢与偏见》英文原版”),不要直接传书名模糊搜索;请先查出该书目标版本的 **10/13 位 ISBN**(如 `9780141439518`)或 **32 位 MD5**,再执行 `da "<ISBN或MD5>" --json`。
- **学术论文**:若用户只给论文标题,先查出对应的 **arXiv 编号**(如 `2310.06825`)或 **开放获取 DOI**,再执行 `da "<arXiv编号或DOI>" --json`。
- **开源商用字体**:若用户只给字体名称(如“帮我下载霞鹜文楷”、“思源黑体”、“阿里巴巴普惠体”或“JetBrains Mono”),请先查阅 [references/fonts.md](references/fonts.md) 中的实测梯队速查表,优先匹配 **Tier 1 首选源直链**(注意:`> 20 MB` 的全量中文字体严禁走 jsDelivr,必须使用 GitHub Releases 或 `gh-proxy.com`;猫啃网/`100font` 详情页非直链),再调用 `da "<字体直链>" --json` 下载并自动解压。
- **公开课课件与教学视频(含 `not_for` 边界)**:详见 [references/education.md](references/education.md)。B站体系课支持直接传入带 `?p=N` 的分 P 链接精准提取单集视频、`-a` 纯音轨与 `.srt` 字幕;MIT OCW(`ocw.mit.edu`)与高校 `.edu` 开放课件支持提取讲义 `.pdf` 并自动生成 `citation.bib`。**严守 `not_for` 边界**:严禁向用户承诺或尝试抓取 Coursera、edX、Udemy、中国大学MOOC(`icourse163.org`)、学堂在线及 B站付费课堂(`bilibili.com/cheese`)等受登录墙或 DRM 加密保护的商业网课。
- **应用安装包**:若用户只给应用名称,先查出对应的 **Android 包名**(如 `org.videolan.vlc`)或 **Microsoft Store 12 位 Product ID**(如 `9N0DX20HK701`),再交给 `da`。
- **GitHub 仓库与目录**:若用户需要克隆整个仓库或拉取多文件目录,优先使用环境中已有的 `gh` CLI、`git` 或 GitHub MCP;仅单文件(`github.com/.../blob/...`)或 Release 发行包直链(`github.com/.../releases/download/...`)使用 `da`。
### 第三步:大文件预检与 Agent 命令超时保护(防卡死必读)
长视频(如 4K 视频)、网盘大合集、安装包体积可能达到数百 MB 至数 GB,下载传输耗时极易超过 Agent 终端命令的默认等待超时(如 60s 或 120s)导致进程被强杀:
1. **优先使用 `--check` 预检**:面对网盘分享、高清视频或未知大小的资源,先执行 `da "<目标>" --check --json` 探测元数据与 `expected_size`(百度/夸克预检不触发云端转存)。
2. **动态调整超时与单文件上限**:
- 若预检体积(`expected_size`)或实际文件超过默认 500 MB,主动追加 `--max-mb <容量>`(如 `--max-mb 2048`)。
- 面对大文件下载时,Agent 调用终端工具应将超时阈值(如 `timeout` 或 `WaitMsBeforeAsync`)调大至充足时间,避免过早超时中断。
3. **异步/后台执行模式**:在支持后台任务(Daemon / Background Task)的 Agent 框架中,GB 级大文件任务建议置于后台执行,稍后轮询退出状态与交付结果。
### 第四步:执行下载命令
**自动化调用始终附加 `--json`**,从 `stdout` 解析结构化 JSON:
```bash
# 通用下载(直接传入分享口令、URL、DOI、ISBN、MD5、包名/Product ID 或本地音频路径)
da "<目标输入>" --json
# 仅提取纯音轨(.m4a / .mp3)
da "<视频或播客链接>" -a --json
# 归档至本地 Obsidian 知识库
da "<公众号或文章链接>" --vault "/path/to/vault" --json
# 仅预检目标标题、大小与元数据(网盘预检不会触发云端转存或创建目录)
da "<目标链接>" --check --json
# 捕获或导入本机浏览器/客户端的目标平台登录 Cookie
da login <平台名称> --json
```
可选参数:`-o <目录>` 指定输出目录,`-p <密码>` 指定提取码或解压密码,`-f` 忽略缓存强制重下,`--max-mb N` 调整单文件大小上限(默认 500 MB),`--no-unzip` 保留原压缩包不自动解压,`--no-sidecar` 不生成额外配套文件。
### 第五步:消费结构化归档产物(赋能下游分析)
下载完成后,**不要直接读取 `.mp4` / `.m4a` 等二进制媒体文件**,应直接利用 JSON 返回的结构化文本文件完成用户的总结、提炼或问答需求:
- **总结视频 / 播客**:直接读取 JSON 中 `sidecars.subtitle`(`.srt` 字幕)或 `sidecars.markdown`(节目 Shownotes、视频简介、时间戳与互动数据)。
- **精读公众号 / 技术文章**:直接读取 `primary_asset`(`article.md`),正文内插图已转为同目录下的本地图片文件。
- **整理学术文献**:读取 `primary_asset`(`.pdf`)的同时,可直接读取 `companion_files` 中的 `citation.bib` 获取标准 BibTeX 引用,以及 `sidecars.markdown` 获取摘要与作者元数据。
## 3. 状态判定与异常自愈 SOP
解析 `--json` 输出的 `status` 与退出码,如实反馈并按规则自愈:
- **`status: "success"`(退出码 `0`)**:主文件已通过格式/媒体结构校验(见 `verification.details.level`),所有配套文件与同步步骤均已完成。
- **`status: "cached"`(退出码 `0`)**:命中本地 SQLite 缓存(`~/.download-anything/index.db`)且本地文件完整,直接使用返回的 `primary_asset` 与 `sidecars` 路径,无需重复下载(除非用户要求强制刷新才加 `-f`)。
- **`status: "partial"`(退出码 `2`)**:**主文件已成功下载并通过校验**,仅部分附属文件(如某张配图、字幕 429)或后处理步骤(如 Obsidian 同步、临时目录清理)未完成。
- **处理准则**:**切勿将 `partial` 当作失败去盲目重下主文件!** 直接使用已落盘的 `primary_asset`,并读取 `warnings` 与 `operations` 字段向用户说明哪项附属步骤未完成及其原因。
- **`status: "dry_run"`(退出码 `0`)**:仅完成 `--check` 预检,未实际落盘文件。
- **`status: "error"`(退出码 `1`)**:根据 `error_type` 执行有限自愈,严禁把失败或预检汇报为下载成功:
- `AuthenticationRequiredError`:目标平台(如百度网盘、夸克网盘、小红书视频、Instagram 多图)需要登录凭据。
- **本地桌面环境**:运行 `da login <平台> --json` 自动提取本机浏览器已登录会话,或提示用户在浏览器中先登录该站点。
- **无头(Headless)环境 / 容器 / 远程服务器**:百度和夸克可注入 `DA_<平台>_COOKIE` 或 `<平台>_COOKIE`,或配置权限为 `0600` 的 `~/.download-anything/credentials.yaml`;Instagram 多图需使用 `DA_INSTAGRAM_COOKIE_FILE` 指向授权 Cookie 文件。
- **安全红线**:严禁在输出或对话中打印会话 Cookie、登录凭据或密码明文。
- `FileTooLargeError`:超过默认 500 MB 上限。若用户明确需要下载该大文件,调高 `--max-mb N` 重试一次。
- `ExtractionError`(含压缩包格式限制):自动解压仅支持受限的 `ZIP` / `TAR`;若因 `7z`、`RAAsk, share a tip or report that it does not work. Sorted by what people marked helpful.
No posts yet. Be the first to ask or share.
Posts here are written by users, not the PopNx team.