AI Assistants (MCP)
Connect Your AI Assistant
Connect your AI assistant to Tutorial AI and ask it, in plain language, to work on your projects: read and rewrite scripts, edit slides, translate, comment, and render or publish videos. Your assistant works as you, with the same access you have in Tutorial AI.
The connection uses MCP (Model Context Protocol), the standard AI assistants use to connect to other apps. You add Tutorial AI to your assistant once, approve access, and you’re ready.
Supported assistants: ChatGPT, Claude, Cursor, Grok, and Gemini.
Before you start
Section titled “Before you start”- A paid Tutorial AI plan
- An account with your AI assistant that can add custom connectors (see the notes for each assistant below)
- About two minutes
You don’t need an API key, and you never paste your Tutorial AI password into your assistant.
Step 1: Copy your MCP URL
Section titled “Step 1: Copy your MCP URL”-
In Tutorial AI, open Settings.
-
Under User Settings, open the Connected apps tab.
-
Next to MCP URL, click the copy button. The URL is:
https://api.videoscripter.ai/api/mcp
Connected apps is a personal setting. Each person on your team connects their own assistant, and it only reaches what that person can reach.
Step 2: Add Tutorial AI to your assistant
Section titled “Step 2: Add Tutorial AI to your assistant”Pick your assistant. Menu names in AI assistants change often. If an option has moved, look for Connectors, Custom connector, or MCP server in your assistant’s settings.
- In ChatGPT on the web, open Settings → Apps & Connectors.
- Open Advanced settings and turn on Developer mode.
- Click Create to add a new connector.
- Fill in the form:
- Name:
Tutorial AI - Description: optional, for example
My Tutorial AI projects - MCP server URL: paste your MCP URL
- Authentication: OAuth
- Name:
- Confirm that you trust the connector, then click Create. A Tutorial AI page opens in your browser. Continue with Step 3.
- To use it, start a new chat, open the + menu, and turn on Tutorial AI under Developer mode.
- In Claude on the web or in the desktop app, open Settings → Connectors.
- Click Add custom connector.
- Enter:
- Name:
Tutorial AI - Remote MCP server URL: paste your MCP URL
- Name:
- Click Add, then Connect. A Tutorial AI page opens in your browser. Continue with Step 3.
- To use it, open the tools menu in a chat and make sure Tutorial AI is turned on.
Using Claude Code? Run this in your terminal, then run /mcp inside Claude Code and choose Tutorial AI to authorize:
claude mcp add --transport http tutorial-ai https://api.videoscripter.ai/api/mcp-
Open Cursor Settings → Tools & MCP.
-
Click New MCP Server. Cursor opens your
mcp.jsonfile. -
Add Tutorial AI and save the file:
{"mcpServers": {"tutorial-ai": {"url": "https://api.videoscripter.ai/api/mcp"}}}If the file already lists other servers, add the
tutorial-aientry inside the existingmcpServersblock. -
Back in Tools & MCP, click Connect next to tutorial-ai. A Tutorial AI page opens in your browser. Continue with Step 3.
-
When the server shows a green dot, it’s ready. Ask for Tutorial AI tasks in Agent mode.
- In Grok on the web, open Settings → Connectors.
- Click New connector and choose Custom.
- Enter
Tutorial AIas the name and paste your MCP URL. - Save, then click Connect. A Tutorial AI page opens in your browser. Continue with Step 3.
- In a chat, make sure Tutorial AI is turned on in the connectors menu.
- In Gemini on the web, open Settings → Connected apps.
- Under Custom apps, click Add a custom app.
- Paste your MCP URL and click Next.
- A Tutorial AI page opens in your browser. Continue with Step 3.
- Mention Tutorial AI in your request, or pick it from the apps menu, to use it in a chat.
Using Gemini CLI? Run this in your terminal, then start gemini and run /mcp auth tutorial-ai to authorize:
gemini mcp add --transport http tutorial-ai https://api.videoscripter.ai/api/mcpStep 3: Authorize access
Section titled “Step 3: Authorize access”Your assistant opens a Tutorial AI page in your browser. Sign in to Tutorial AI if asked.
-
Check the heading: <App name> wants to access your Tutorial AI account.
-
Check Signed in as. If it shows the wrong email, sign out of Tutorial AI in this browser and start again from your assistant.
-
Review what the assistant will be able to do:
Permission What it covers View your workspaces and projects Access projects, scripts, slides, comments, settings and assets. View your video content Access narration, on-screen text, markers and source videos. Edit your projects Change scripts, slides, settings and voices; translate projects, add comments, generate videos and manage public access. The screen also lists what it cannot do: record or delete videos, invite users, or change billing.
-
Click Authorize access. Your browser returns to your assistant, and the connection is ready.
A few things to know:
- You see this screen every time you connect. Tutorial AI never grants access silently.
- The link expires after about 10 minutes. If it expires, start the connection again from your assistant.
- Changed your mind? Click Cancel. No access is granted.
Step 4: Try it
Section titled “Step 4: Try it”Ask your assistant something about one of your projects:
- Read the script of Onboarding walkthrough
- Summarise the narration of Onboarding walkthrough for a changelog
- Tighten the script of Onboarding walkthrough and show me the changes before saving
Your assistant picks the right Tutorial AI action for each request. You don’t need to know any commands.
Your new connection also appears in Tutorial AI under Settings → Connected apps, with a New badge for its first 24 hours.