MCP YouTube Transcript
An MCP server retrieving transcripts of YouTube videos
作者提供的说明(英文原文)
Search 4M+ podcasts & YouTube, transcribe any episode, search transcripts, generate AI lessons.
这三项永远不会合并成一个分数:意见、测量和人气是不同的东西。
我们还没有为它写通俗的介绍。下面是我们从注册中心读取到的事实。
选择你使用的 AI 应用,然后复制设置。
Authorization标为机密的值是你账户的钥匙。千万不要分享给别人,也不要粘贴到聊天里。
你需要
你用的是哪个应用?
在上方选择你的应用。我们会记住它,下次按钮会显示“安装到 …”。
这个地址需要在请求的 header 中放入密钥,而添加连接器的菜单无法设置。请改用 Claude Code、Cursor 或 VS Code。
在终端中运行这条命令
claude mcp add --transport http podlearn https://podlearn.ai/mcp --header 'Authorization: your-value'运行前请先填写空着的值。
粘贴到项目中的 .cursor/mcp.json,或 ~/.cursor/mcp.json
{
"mcpServers": {
"podlearn": {
"url": "https://podlearn.ai/mcp",
"headers": {
"Authorization": "your-value"
}
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
粘贴到项目中的 .vscode/mcp.json
{
"servers": {
"podlearn": {
"type": "http",
"url": "https://podlearn.ai/mcp",
"headers": {
"Authorization": "your-value"
}
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
这个地址需要在请求的 header 中放入密钥,而添加连接器的菜单无法设置。请改用 Claude Code、Cursor 或 VS Code。
你的选择通过 Cookie 保存在这台设备上,无需账户。
Authorization 机密 HTTP header这些设置是根据作者在注册中心的条目生成的。如果 README 的写法不同,请以 README 为准。
我们还没有读取这个服务器的工具列表;我们会轮流检查各个端点。
我们检查的是任何人都能看到的信息;我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月7日
注册名称已通过 GitHub 账户 blutrich 验证,该账户拥有此仓库
我们还没有从 GitHub 读取这个仓库
我们还没有从 GitHub 读取这个仓库
尚未检查;我们会轮流检查各个端点
尚未检查;我们会轮流检查各个端点
尚未检查;我们会在检查端点时读取工具列表
与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时,星标数才计入。
它在 1 项适用的基础检查中通过了 1 项。我们没有阅读代码。在添加密钥或令牌之前,请先看看这个服务器要求哪些权限,并使用只具备所需权限的密钥。
你需要填写:Authorization
它在作者的云端运行。通过互联网连接,无需安装,但你发送的内容会经过作者的服务器。
GitHub 仓库最后更新于–,最新版本(1.1.0)于19天前进入注册中心。
支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。它通过 URL 连接,所以接受远程 MCP 服务器的应用也可以使用,例如开启开发者模式的 ChatGPT。
同一分类中有:MCP YouTube Transcript, Google Analytics MCP, BibiGPT YouTube Video Summarizer
同一分类中其他作者的服务器,按星标数排序。
如何看卡片:★ GitHub 星标i三大支柱:评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i
An MCP server retrieving transcripts of YouTube videos
GA4 MCP server for AI agents: schema discovery, server-side aggregation, analysis-ready defaults.
AI video summarizer: summarize and transcribe YouTube, Bilibili, TikTok and podcast videos.
按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。
The podcast & video knowledge engine, right inside Claude Code. Drop a link. Get a verbatim transcript, a searchable archive, and an AI lesson in seconds.
Some of the best thinking on the planet is locked inside hours of audio you'll never have time for. PodLearn unlocks it. Hand Claude any podcast or YouTube URL and it transcribes the whole thing (yes, even long talks with no captions), makes every word searchable, and distills the ideas into a structured lesson. No tab-switching, no copy-paste, no "I'll get to it later."
One install adds the PodLearn **skill *and* MCP server** to Claude Code. The skill handles the routing; the MCP does the work. You just talk.
transcribe_feed ingests an entire podcast feed in one shot, not episode-by-episode.*"What podcasts do you have about AI agents?"*
*"Transcribe this and make me a lesson: ‹youtube or podcast URL›"*
*"Find the best quotes about pricing from ‹episode›"*
*"Transcribe the last 10 episodes of ‹podcast› and summarize the themes."*
*"Turn that lesson into a LinkedIn post."*
1. Have a PodLearn account (free, 2 credits to try) at [podlearn.ai](https://podlearn.ai). No token needed: you sign in once in your browser.
2. Add the marketplace — in Claude Code (run this on its own, wait for confirmation):
/plugin marketplace add blutrich/podlearn-plugin
3. Install the plugin — then run:
/plugin install podlearn@podlearn
⚠️ Run them one at a time — they're two separate slash commands. Pasting
both lines at once makes Claude treat the second as part of the first.
4. Restart Claude Code (or claude reload).
5. Sign in. Run /mcp, pick podlearn, choose Authenticate, and approve PodLearn in the browser. Approve both read and write if you want Claude to transcribe and save things for you (it still asks before spending credits). Then start asking.
That's it. The plugin ships the skill and the MCP, so there's no separate claude mcp add step and nothing to wire up by hand.
Skill (podlearn) | Routes your intent through the MCP tools so you never think about which one to call. |
| MCP server | search_podcasts · search_youtube · search_episodes · list_feed_episodes · transcribe_url · transcribe_feed (bulk/season) · get_transcription · search_transcription · generate_lesson · generate_linkedin_post · and more. |
The plugin already bundles the server, so most people don't need this. But if you want the MCP on its own — in Claude Desktop, Cursor, ChatGPT, or any other MCP client, or in Claude Code without the skill — add it directly.
Claude app (web, desktop, mobile): Settings → Connectors → Add custom connector → paste https://podlearn.ai/mcp → Connect, and sign in.
Claude Code:
claude mcp add --transport http podlearn https://podlearn.ai/mcp
Then /mcp → podlearn → Authenticate.
Codex:
codex mcp add podlearn --url https://podlearn.ai/mcp
codex mcp login podlearn --scopes mcp:read,mcp:write
Any MCP client — point it at the Streamable HTTP endpoint and sign in with OAuth:
| Field | Value |
|---|---|
| Transport | Streamable HTTP (http) |
| URL | https://podlearn.ai/mcp |
| Auth | OAuth (sign in with your PodLearn account) |
Headless client that can't open a browser? Create a personal token in Settings → API Tokens and send it as the Authorization: Bearer … header. Never put it in the URL: tokens in the query string are ignored, and calls will hang waiting for a sign-in. Full setup guide: podlearn.ai/mcp-docs.
/plugin marketplace update podlearn # pull the latest
/plugin uninstall podlearn@podlearn # remove the plugin /plugin marketplace remove podlearn # remove the marketplace
/mcp → podlearn → re-authenticate. To get write access after a read-only sign-in, remove and re-add the server first./mcp. You can revoke the old pl_mcp_… token in Settings → API Tokens.transcribe_url, transcribe_feed, generate_lesson) spend PodLearn credits, and the skill confirms the cost before charging. Start with 2 free credits; top up or subscribe at podlearn.ai.PodLearn runs as a web app at [podlearn.ai](https://podlearn.ai) and as an MCP server for Claude Desktop, ChatGPT, Cursor, and any MCP-compatible client. This plugin is the fastest way to use it from Claude Code.
AGPL-3.0 — © 2026 Ofer Blutrich. Built on the PodLearn platform.