ChatGPT MCP Connector

You can now connect your EmbedSocial account to ChatGPT using MCP (Model Context Protocol). Just name a social account, widget, or source, and ChatGPT can pull your content, analytics, or media directly. No exporting, no copy-pasting. 🤖

Prefer Claude? The same MCP server works there too, see Connect EmbedSocial to Claude with MCP.

What is MCP?

MCP is an open standard that lets AI assistants securely connect to external tools and data. EmbedSocial provides an MCP server, so once you connect it, ChatGPT can read your EmbedSocial content and analytics on your behalf, using your existing account and permissions.

Think of it as a secure bridge: ChatGPT can access the same content and metrics you can see in EmbedSocial, without you having to export, paste, or switch between tools.

What you can do once it's connected

Give ChatGPT the name of a connected social account and it can pull that account's posts, UGC, and reviews. For engagement analytics (post, icon, lightbox, and CTA clicks), name the widget instead. Your media library works differently still: name the source it should come from. With it connected, ChatGPT can:

• Pull the social posts, UGC, and reviews from a social account you name
• Filter that content by date range or by keyword
• Look up an individual post or review from that account
• Read a widget's engagement analytics, post, icon, lightbox, and CTA click activity over a chosen period
• Summarize, theme, and turn that content into reports, recaps, and drafts right inside the chat
• Retrieve your media by naming the source it came from

ChatGPT only reads what's already in your account. It works through your EmbedSocial API key, so it can access the same data you can, and nothing more.

How to connect your EmbedSocial custom connector

Integrate EmbedSocial as a custom connector to analyze your social media data directly. These steps cover setup, connection, and activation.

Custom connectors in ChatGPT require Developer mode to be turned on first. You'll only need to do this once.

In ChatGPT, open Settings, go to Security (under Security and login), and enable Developer mode. This is required before you can add any custom connector.

Go to the Plugins section and click the plus (+) button to add a custom plugin.

Enter "EmbedSocial" in the name field and provide our specific Remote MCP server URL required for your configuration: https://embedsocial.com/admin/mcp. Then click Create.

You'll be guided to sign in with your EmbedSocial account to authenticate the connection and enable tool permissions.

Once connected, use the chat to request data analysis. ChatGPT will now access your EmbedSocial tools to process your social reviews.

Use cases & example prompts

• "How many CTA clicks did my 'Homepage Reviews' widget get in the last 30 days?"
• "Show me the post, lightbox, and CTA clicks for my 'Summer Campaign' widget over the last 90 days."
• "Which interaction is highest on my 'Instagram Feed' widget in the last 7 days, posts, icon, lightbox, or CTA?"
• "Draft a short performance recap for my 'Testimonials' widget based on its last-30-day analytics."

• "Pull the data from my 'Trustpilot' account and draft a short recap I can share with my manager."
• "Write three social captions based on the best reviews in my 'Google Business Profile' account."
• "Summarize this month's feedback from my 'Yelp' account into themes for the product team."

• "Get my recent media from 'Instagram' so I can pick a few images for next week's post."
• "Show me my most recent uploaded images from 'Google'."

What data can the AI access?

The connection is read access to your own EmbedSocial content. Name a social account for posts, UGC, and reviews; name a widget for engagement analytics; name a source for media. Either way, it uses your account's API key, so ChatGPT sees the same data you do and can't reach anything outside your account.

If you have any questions, click the bottom right chat icon to talk with our customer success team or send us an email at support@embedsocial.com We're here for you 🙏