MCP YouTube Transcript
An MCP server retrieving transcripts of YouTube videos
作者提供的说明(英文原文)
Query your SEO data in plain language: rankings, audits, backlinks, competitors and AI visibility.
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。
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 …”。
适用于网页版、桌面版和手机版 Claude:
https://app.miraqo.io/mcp在终端中运行这条命令
claude mcp add --transport http seo https://app.miraqo.io/mcp粘贴到项目中的 .cursor/mcp.json,或 ~/.cursor/mcp.json
{
"mcpServers": {
"seo": {
"url": "https://app.miraqo.io/mcp"
}
}
}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": {
"seo": {
"type": "http",
"url": "https://app.miraqo.io/mcp"
}
}
}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。
ChatGPT 只能连接互联网上的服务器:
https://app.miraqo.io/mcp菜单名称可能因套餐和版本不同而略有差异。
你的选择通过 Cookie 保存在这台设备上,无需账户。
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6日
发布者(deleteweb)不是仓库所有者(Miraqo-io);可能是其他人代为上架的
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 0 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。
注册中心没有列出必需的密钥,不过有些服务器会在首次使用时要求你登录。
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(1.0.0)于3个月前进入注册中心。
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。
Connect Claude, ChatGPT, Gemini, Claude Code or Cursor to your SEO data in Miraqo — rankings, technical audit, backlinks, competitors, Search Console and AI visibility — and query it in plain language from the chat you already use.
This is a hosted, remote MCP server. There is nothing to install and nothing to run locally. This repository is documentation only.
https://app.miraqo.io/mcp
You need a Miraqo account. A 14-day free trial is available, no card required.
Add a custom connector and paste the address above. A standard OAuth authorization starts automatically: sign in to Miraqo, confirm, done. No token to generate or paste.
Where you add a connector differs slightly between apps (and may depend on your plan with that app), but the address is always the same.
You can revoke access at any time from Account settings → AI Connector (MCP) → Connected apps.
Generate a personal token in Account settings → AI Connector (MCP), then:
claude mcp add --transport http miraqo https://app.miraqo.io/mcp \
--header "Authorization: Bearer YOUR_TOKEN"
{
"mcpServers": { "miraqo": { "url": "https://app.miraqo.io/mcp", "headers": { "Authorization": "Bearer YOUR_TOKEN" } } } }
Tokens are shown once at creation. You can keep up to 5 active and revoke them individually.
26 tools. Start with list_projects to get the project_id the others take.
| Tool | What it returns |
|---|---|
list_projects | Your projects, each with its latest audit (id, status, date, score) |
get_usage | Month-to-date consumption against plan limits, including the MCP action cap |
get_keywords | Tracked keywords with latest position, URL and AI Overview presence |
get_keyword_history | Up to 365 days of position history for one keyword, with SERP features |
get_competitors | Tracked competitors: average position and keyword-gap size |
get_keyword_gap | Keywords a competitor ranks for, with volume, difficulty and CPC |
get_domain_analysis | Cached competitive snapshot of any domain: authority, organic keywords, top pages |
get_backlinks | Latest backlink snapshot: new, lost, referring domains, anchors, spam score |
get_audit_overview | Score, pages crawled and what the latest audit produced |
get_audit_issues | Non-zero issue counters: 404s, titles, duplicates, security, sitemap |
get_audit_pages | Per-page status, title, word count, load/LCP/TTI and issue flags |
get_audit_redirects | Redirect chains with linked_from — the pages where the fix belongs |
get_cwv_report | Core Web Vitals: lab and field data, worst pages by LCP |
get_topical_architecture | Topic clusters with their pillar pages, structural or AI-refined |
get_cluster_pages | Pages in a cluster, ranked by inbound internal links |
find_pages | Search crawled pages by URL or title substring |
get_ai_visibility | Whether AI assistants mention your brand and cite your URLs, with share of voice |
get_gsc_performance | Search Console by query or page, aggregated over 7/28/90-day windows |
get_gsc_daily | Day-by-day Search Console site totals — the series to date a drop against |
Each also counts against a separate monthly cap on MCP-initiated actions. Check get_usage before running them.
| Tool | What it does |
|---|---|
run_audit | Starts a new technical crawl. Costs audit credits equal to the pages actually analysed |
run_keyword_research | Volume, difficulty and CPC — by seed, list, domain, or gap against a competitor |
run_domain_analysis | Refreshes a competitive snapshot. Reuses a snapshot newer than 7 days for free |
run_ai_visibility_check | Re-runs the project's AI visibility prompts |
add_ai_prompt | Creates an AI visibility prompt and runs the first check |
refine_topical_architecture | AI pass over the audit's topic clusters. Once per audit |
generate_topical_bridge | Writes a linking paragraph to connect two pages. Cached pairs are free |
Reads are unmetered, so exploring your data costs nothing.
The same capabilities are available as a documented REST API under /api/v1, for integrations that are not MCP clients. The link is in the AI Connector card in your account settings.
Issues and questions about the connector are welcome here. For account or billing matters, write to supporto@miraqo.io.