Skip to main content
TubeKit AI · MCP

Ask your AI.
Answered by TubeKit.

Plug TubeKit into Claude, ChatGPT, Cursor or Codex and let your assistant pull real channel, competitor and keyword data — right inside the chat.

Compatible with
Claude ChatGPT Cursor Codex
Compare my channel growth against my top 3 competitors.
competitor_analysis · 3 Credits
MrBeastKurzgesagtVeritasiumYou
AI
Claude · ChatGPT · Cursor
MCP
TubeKit MCP
TubeKit
Credits
Why TubeKit MCP

Not a generic API — your TubeKit engine

The same creator data that powers the TubeKit dashboard — now inside your AI tools.

One TubeKit account

Your API key links your TubeKit plan. No separate billing, no extra sign-ups.

Creator-native data

Channel analysis, competitor analysis and keyword research — the same engine the dashboard uses.

60-second setup

Paste a URL, add your API key. No config files to debug, no docs to read.

Secure by default

API keys are hashed, scoped to MCP only, and revocable anytime. Read-only — never touches your channel.

Natural language

Ask like you would ask a colleague. No query syntax, no API calls to learn.

Works with everything

MCP is an open standard. Claude, ChatGPT, Cursor, Codex — if it speaks MCP, it works.

Setup guide

Connect in 60 seconds

Pick your AI tool and follow the steps. Your server URL and API key come from your TubeKit dashboard.

https://tubekitai.com/mcp

You will get your personal API key from the MCP page in your dashboard after signing in.

  1. 1 Open Claude.ai → Settings → Connectors → Add custom connector.
  2. 2 Select MCP Server and paste the Server URL above.
  3. 3 Click Connect, then paste your API key to authorize access.
  4. 4 Done — TubeKit tools are now available in your conversations.
  1. 1 Open ChatGPT → Settings → Connectors and choose Add custom MCP server.
  2. 2 Paste the Server URL above.
  3. 3 Authenticate with your API key.
  4. 4 Start a new chat and ask TubeKit anything about your channel.
  1. 1 Open Cursor → Settings → MCP → Add new server.
  2. 2 Paste the Server URL and add your API key as a Bearer token.
  3. 3 Enable the server.
  4. 4 Use TubeKit tools directly from Cursor chat.
  1. 1 Open your Codex MCP configuration.
  2. 2 Add TubeKit with the Server URL and header Authorization: Bearer <your API key>.
  3. 3 Reload Codex.
  4. 4 TubeKit tools are now available to the agent.
Claude

Connect in Claude — one click

Add the connector once, then sign in with your TubeKit AI account. No API key needed.

1
Open Settings and click Customize.
Open Settings and click Customize.
2
Go to Connectors and click Add (top right).
Go to Connectors and click Add (top right).
3
Name it, paste the server URL and click Continue.
Name it, paste the server URL and click Continue.
4
Keep "register one automatically" selected and click Add.
Keep "register one automatically" selected and click Add.
5

Click Connect, sign in with your TubeKit AI account and approve — done.

Steal these prompts

Paste these straight into your AI tool once TubeKit is connected.

Competition

Who are the fastest-growing channels in my niche this month?

Competition

Compare my channel growth against my top 3 competitors.

Performance

Why did my last 5 videos underperform? Show me the data.

Performance

What is my average engagement vs. my niche benchmark?

Keywords

Find low-competition keywords for a beginner [topic] video.

Keywords

What long-tail keywords should I target for [topic]?

Analytics

Give me a channel health report with key insights.

Analytics

Analyze this channel and tell me what is working: [channel URL].

Tools you can call

Ask in plain language — the AI picks the right tool. Paid tools spend credits from your plan.

channel_analysis

Analyses a YouTube channel: subscriber and view metrics, upload cadence, performance score and an AI read. Accepts a URL or an @handle.

1 credit
competitor_analysis

Compares your channel with 1–5 competitors: content strategy, titles and thumbnails, upload cadence and where your edge is.

3 credits
keyword_research

Returns related searches, long-tail suggestions, an interest curve, YouTube competition and a difficulty score for a keyword. Free.

Free
video_analysis

Analyses a single YouTube video: metrics, SEO and an AI performance read with concrete improvements.

1 credit
comment_analysis

Analyses a video’s comments: sentiment split, recurring themes, frequent questions and a viewer feedback summary.

1 credit
niche_analysis

Suggests viable YouTube niches from your interests: audience size, competition, monetisation potential and a way in.

2 credits
video_ideas

Generates 10 click-focused video ideas for a topic: title, hook line and why each one should work.

1 credit
seo_package

Produces an SEO package for a video topic: optimised title options, description, tags and thumbnail text ideas.

1 credit
keyword_trends

Analyses current YouTube trends for a niche: rising subtopics, formats that are working and angles worth trying.

1 credit
generate_thumbnail

Generates a click-focused YouTube thumbnail from a text prompt and returns its link.

2 credits
brand_asset

Generates a brand asset for your channel (logo, banner, poster and so on) and returns its link.

2 credits
gap_finder

Finds unserved but in-demand content gaps for a channel or niche: topic, why the demand exists, a suggested title and difficulty.

3 credits
credits_balance

Returns the current credit balance on your account. Free helper tool.

Free

Common questions

MCP (Model Context Protocol) is an open standard that lets AI tools connect to external data. TubeKit’s MCP server gives Claude, ChatGPT, Cursor, Codex and others direct access to TubeKit’s YouTube tools — so you ask in plain English and get real, data-backed answers.

Any TubeKit account works. MCP tool calls spend credits from your plan exactly like the dashboard does; free tools (like keyword research and balance checks) cost nothing.

One-click connect (OAuth) where your client supports it — that is the shortest path. Otherwise generate an API key on the MCP page in your dashboard, paste it into your AI tool, and revoke it anytime. Keys are scoped to MCP only.

Yes — paid tools (channel analysis, competitor analysis) spend the same credits as the website. You always see the cost per tool, and a failed call is never charged.

No. TubeKit MCP tools are read and analysis only. They never post, edit or delete anything on your channel.

Any MCP-compatible client: Claude, ChatGPT, Cursor, Codex and more. MCP is an open standard, so support keeps growing.

API keys are stored hashed and shown only once. All traffic is authenticated over HTTPS, and you can revoke any key instantly.

Bring TubeKit into your AI tool

Create a free account, grab your API key, and start asking.