← 返回 Skills 市场
YouTube/B站 视频搜索下载
作者
joppyao-bit
· GitHub ↗
· v1.0.0
· MIT-0
153
总下载
0
收藏
4
当前安装
1
版本数
在 OpenClaw 中安装
/install yt-search-download
功能描述
多站点视频搜索、下载、字幕提取工具。支持 YouTube、B站（Bilibili）等主流平台。 结合 YouTube Data API v3 进行高级搜索，yt-dlp 下载视频/音频/字幕。 核心能力：全站关键词搜索、频道浏览、按时间/播放量/相关度排序、下载视频、提取音频（MP3）、下载字幕（中英文）、查看视...
安全使用建议
Before installing: (1) realize the skill requires a YouTube API key (set YT_BROWSE_API_KEY or YOUTUBE_API_KEY) — the registry metadata omitted this, so it may fail if you don't provide it; (2) the README claims Bilibili support but the included script only implements YouTube search (you can still use yt-dlp to download Bilibili URLs, but channel/search features for Bilibili are not implemented); (3) the tool runs yt-dlp and recommends --cookies-from-browser, which can read browser cookies/profiles — only allow that if you trust the code and understand the privacy implications; (4) the skill will write downloads and generated .srt/.txt files to your Downloads directory; (5) if you care about safety, review scripts/yt_search.py locally (it uses subprocess.run without shell=True and uses the official Google Data API endpoints, which is good) and confirm you are comfortable with the forced auto-translation behavior specified in SKILL.md. If anything above is unexpected, don't install or run the skill until the author corrects the metadata and clarifies Bilibili support and cookie usage.
功能分析
Type: OpenClaw Skill
Name: yt-search-download
Version: 1.0.0
The skill bundle provides legitimate functionality for searching and downloading YouTube and Bilibili content using the YouTube Data API v3 and the yt-dlp utility. The Python script (yt_search.py) and instructions (SKILL.md) are well-documented and align with the stated purpose of video search, metadata retrieval, and subtitle extraction. While the skill includes instructions for the AI to access browser cookies (via yt-dlp) and execute a Python one-liner for subtitle processing, these actions are transparently presented as necessary for bypassing video restrictions and formatting data for the agent. No evidence of data exfiltration, malicious intent, or harmful prompt injection was found.
能力评估
Purpose & Capability
The description advertises multi-site support (YouTube and Bilibili) and advanced search, and the code and SKILL.md clearly implement YouTube Data API search + yt-dlp downloads. However: (1) the registry metadata lists no required environment variables even though the tool needs an API key (YT_BROWSE_API_KEY or YOUTUBE_API_KEY); (2) Bilibili search/browsing capability is claimed in README/SKILL.md but the provided script only calls YouTube APIs and formats YouTube URLs (no Bilibili API/search implementation). These are functional mismatches that could surprise users.
Instruction Scope
SKILL.md prescribes using the script at ~/.claude/skills/… and running yt-dlp commands (including --cookies-from-browser chrome) and auto-translating titles without asking the user. The download/subtitle instructions will cause the agent to: call Google APIs, run yt-dlp (which may access browser cookies), write .srt/.txt files to ~/Downloads, and post-process SRT into TXT via a small Python snippet. These actions are consistent with download tasks but the forced translation behavior (must replace '【译】___' without asking) is a policy/scope decision the user should be aware of.
Install Mechanism
No install spec is provided (instruction-only skill with included script file). No external archives or untrusted downloads are specified. The only runtime dependency (yt-dlp) is a common third‑party tool the SKILL.md asks the user to install via brew/pip — no hidden installer URLs or extract steps present.
Credentials
The code requires a YouTube API key (reads YT_BROWSE_API_KEY or YOUTUBE_API_KEY) but the registry metadata declared no required env vars — an inconsistency. Besides that, the skill does not request other credentials. Note: instructions encourage yt-dlp's --cookies-from-browser which can access browser cookies (sensitive), so users should understand that downloading certain videos may cause access to local browser profile data via yt-dlp.
Persistence & Privilege
The skill does not request always:true and does not modify other skills or system-wide settings. It will store its script under the skills directory (normal). It can be invoked autonomously (default), which is expected for skills, but that is not combined here with elevated privileges or hidden credentials.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install yt-search-download - 安装完成后，直接呼叫该 Skill 的名称或使用
/yt-search-download触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.0.0
yt-search-download 1.0.0 – 多站点视频搜索、下载与字幕工具
- 支持YouTube、B站等主流平台的视频/音频/字幕搜索与下载。
- 提供高级搜索：关键词、频道、时长、播放量、时间范围等多种筛选方式。
- 一键下载视频、提取音频（MP3）、下载字幕（自动输出中英SRT+TXT）。
- 浏览频道、查看视频详情，所有输出按严格格式与中英文译文规范。
- 结合YouTube Data API v3与yt-dlp，支持命令行快速操作与多种下载需求。
元数据
常见问题
YouTube/B站 视频搜索下载 是什么？
多站点视频搜索、下载、字幕提取工具。支持 YouTube、B站（Bilibili）等主流平台。 结合 YouTube Data API v3 进行高级搜索，yt-dlp 下载视频/音频/字幕。 核心能力：全站关键词搜索、频道浏览、按时间/播放量/相关度排序、下载视频、提取音频（MP3）、下载字幕（中英文）、查看视... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 153 次。
如何安装 YouTube/B站 视频搜索下载？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install yt-search-download」即可一键安装，无需额外配置。
YouTube/B站 视频搜索下载 是免费的吗？
是的，YouTube/B站 视频搜索下载 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
YouTube/B站 视频搜索下载 支持哪些平台？
YouTube/B站 视频搜索下载 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 YouTube/B站 视频搜索下载？
由 joppyao-bit（@joppyao-bit）开发并维护，当前版本 v1.0.0。
推荐 Skills