Video Catcher

当用户要求下载视频、保存网页中的视频、下载 mp4/m3u8/mpd、下载 YouTube/Bilibili/抖音/TikTok/X/Instagram 等平台视频、批量下载视频、下载字幕/封面、检查网页中有哪些媒体资源、查看视频有哪些可下载清晰度、指定 360p/720p/1080p/2K/4K 清晰度、继续中断的视频下载时使用。面向 Codex、Claude Code 等本地 Agent:用户只需提供链接和目标,skill 自动执行 yt-dlp、YouTube PO Token Provider、浏览器指纹重试、静态发现、隔离持久浏览器网络嗅探、请求上下文继承、音视频合并、DRM 检查、断点续传、验证与报告;默认不依赖用户正在运行的 Chrome/Edge Cookie 数据库,也不在下载过程中要求用户选择技术路线。

49 stars 11 forksImages, Video & Audio

Install

skills CLI (Claude Code, Cursor, Codex and more)
npx skills add Weikezi-AI/video-catcher --skill video-catcher
Manual install
git clone --depth 1 https://github.com/Weikezi-AI/video-catcher.git /tmp/video-catcher
mkdir -p ~/.claude/skills
cp -r /tmp/video-catcher/. ~/.claude/skills/video-catcher

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

# Video Catcher

让本地 Agent 在用户只说“下载这个视频/网页里的视频”时,自行完成发现、解析、下载、恢复和验证。

## 交互原则

**正常下载流程必须是非交互式的。**

用户只给一个链接时:

```bash
python scripts/video_catcher.py download "URL"
```

然后让脚本自行走完整 fallback chain。

Agent 不应在中途询问:

- “要不要换 Firefox?”
- “要不要关闭 Chrome?”
- “要不要手动导 cookies.txt?”
- “选择方案一还是方案二?”

也不要自动杀掉用户的 Chrome/Edge 进程。

只有所有自动路线都失败后,才在最终结果中说明客观阻塞原因。对确实要求账户登录、人工验证码、会员权限或私有权限的内容,不承诺在全新未认证环境中零人工完成。

## 适用请求

- “帮我下载这个视频”
- “把这个网页里的视频保存下来”
- “下载 B 站 / YouTube / 抖音视频”
- “把这些链接都下载”
- “下载这个 m3u8 / mpd / mp4”
- “看看这个页面里有哪些视频资源”
- “这个视频有哪些清晰度 / 能下 4K 吗”
- “先列出清晰度,我再选”
- “下载 1080p / 最高画质 / 中文字幕 / 封面”
- “继续刚才中断的下载”

如果用户只想分析已经存在的本地视频文件,不走下载流程。

## 合规与安全边界

以下约束属于 Skill 的执行规则,而不是可选提示:

1. 只保存用户有权访问、复制和保存的内容;不要因为 URL 可访问就假定用户拥有转载、再分发或商业使用权。
2. 不实现或指导绕过 Widevine、FairPlay、PlayReady 等 DRM、付费墙、会员/账号权限、访问控制或其他版权技术保护措施。
3. 如果用户明确要求绕过 DRM、盗用凭据、突破无权访问的私有内容或规避付费/权限控制,应停止相应流程,而不是寻找替代绕过路线。
4. Cookie、Authorization、Token 等敏感凭据不得写入公开报告、日志正文、聊天回复或提交仓库。
5. 不把 Cookie、Token 或 Authorization 上传给第三方解析服务。
6. 默认单视频、最高 1080p;只有用户明确要求才下载整个播放列表或最高画质。
7. 不为了读取 Cookie 而关闭、杀死或修改用户日常浏览器进程。
8. 下载结果只代表技术处理成功,不代表内容已获得版权、转载、传播或商业使用授权。
9. 软件许可只覆盖 video-catcher 本身,不改变下载内容原有的版权、商标、肖像、隐私或平台许可状态。
10. 完整项目使用声明见 `DISCLAIMER.md`;该文件用于说明预期用途和边界,不构成法律意见。

## 依赖

首次使用:

```bash
python -m pip install -U -r requirements.txt
python -m playwright install chromium
python scripts/doctor.py
```

完整自动 fallback 推荐具备:

- requests
- yt-dlp >= 2025.05.22
- curl_cffi(yt-dlp 浏览器/TLS impersonation)
- bgutil-ytdlp-pot-provider
Read the full skill on GitHub

Details

Skill name
video-catcher
License
GPL-3.0
Repo updated
9 August 2026

FAQ

What does the Video Catcher skill do?

当用户要求下载视频、保存网页中的视频、下载 mp4/m3u8/mpd、下载 YouTube/Bilibili/抖音/TikTok/X/Instagram 等平台视频、批量下载视频、下载字幕/封面、检查网页中有哪些媒体资源、查看视频有哪些可下载清晰度、指定 360p/720p/1080p/2K/4K 清晰度、继续中断的视频下载时使用。面向 Codex、Claude Code 等本地 Agent:用户只需提供链接和目标,skill 自动执行 yt-dlp、YouTube PO Token Provider、浏览器指纹重试、静态发现、隔离持久浏览器网络嗅探、请求上下文继承、音视频合并、DRM 检查、断点续传、验证与报告;默认不依赖用户正在运行的 Chrome/Edge Cookie 数据库,也不在下载过程中要求用户选择技术路线。

How do I install the Video Catcher skill?

skills CLI (Claude Code, Cursor, Codex and more): npx skills add Weikezi-AI/video-catcher --skill video-catcher. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Video Catcher popular?

It lives in Weikezi-AI/video-catcher, which has 49 GitHub stars and 11 forks.

Related skills

Character Lora

80

KerberosClaw/kc_ai_skills

Use when the user wants to build a consistent-identity LoRA for an original character — defining the character, generating a face/body-consistent multi-angle dataset (via the gpt-image-gen skill for codex image generation), captioning it, doing base-specific homework, training on a chosen base (Pony / Z-Image / others) on a local GPU, and producing a usable LoRA. This skill ORCHESTRATES the end-to-end pipeline and gates every expensive/irreversible step; it delegates actual image generation to gpt-image-gen and never improvises training settings from memory.

Images, Video & AudioIn 1 list13

88api Image Gen

49

blackdm666/88API-image-gen

Generate or edit images through 88api.ai with every image model the 88API docs catalog lists — GPT Image (gpt-image-2, gpt-image-2.5-flare/sunburst, az-gpt-image-*), Gemini / Nano Banana (gemini-3.1-flash-image, gemini-nano-banana-2.1, gemini-3-pro-image), Grok Imagine, Seedream and Midjourney. The plugin reads each model's endpoint and parameter limits from the live 88API model catalog and picks the right API (Images API or Chat Completions) itself. Use for text-to-image, reference-image editing, multi-reference composition, 1K/2K/4K output, transparent backgrounds, batches, repeated generati

Images, Video & Audio11

Gpt Image Gen

80

KerberosClaw/kc_ai_skills

Use when the user asks to generate OR edit an image via GPT/Codex (e.g. 「叫 gpt 生圖」「幫我用 gpt 生圖」「gpt 畫一個 X」「幫我去背」「把這張圖的背景去掉」). The skill drafts a Chinese + English prompt pair and waits for explicit approval. After approval it uses the host-native executor: Codex calls its built-in image_gen tool directly, while Claude Code keeps the Codex CLI background workflow. Supports text-to-image, img2img, and precise EDIT mode with preservation constraints and verification where local artifacts are available.

Images, Video & AudioIn 1 list13

Pronounce Word

51

anzy-renlab-ai/pronounce

User asks how to pronounce an English word or project/product name ("how to pronounce X", "pronounce X", "X 怎么读", "X 怎么发音", "读一下 X"). Generate audio via the `say-it` CLI so the user actually HEARS the word — three times by default — instead of only writing IPA/syllable hints. The CLI consults a community-maintained pronunciation dictionary (kubectl → "koob-control", GIF → "jif", JSON → "jay-son", ...) and feeds an English-like respelling to the system TTS engine (macOS `say`, Linux `espeak-ng`, or Windows PowerShell) so project names come out the way engineers actually say them. Triggers on a

Images, Video & AudioIn 1 list

Searxng

80

KerberosClaw/kc_ai_skills

Use when the user wants privacy-respecting web, image, news, or video search through a configured local SearXNG instance instead of external search APIs. Calls the bundled script against SEARXNG_URL, supports result limits/categories/language/time range, and can return human-readable or JSON output. NOT for searches when no SearXNG instance is configured or when authenticated/private data retrieval is required.

Images, Video & AudioIn 1 list13

Cesium Skills

53

OpenCesium/Cesium-Skills

Writes native CesiumJS demos for the latest stable Cesium API (async terrain, 3D Tiles, imagery, WebGL2 shaders/effects). Use when creating or migrating Cesium Viewer, layers, terrain, Cesium3DTileset, models, CustomShader, PostProcessStage, materials, particles, or when old Cesium 1.x / WebGL1 demos break. Do not use for non-Cesium GIS or mars3d/SuperMap wrappers.

Images, Video & Audio

Browse the most popular Claude skills, all images, video & audio skills, or write your own with the skills generator.