Resources

Recommend for Claude: ask your demand data in plain language

Our MCP connector lets an AI assistant read your Recommend data on your behalf: projects, tracked objects, scores and the evidence behind them. It is read-only, scoped to one workspace you choose, and takes a minute to connect.

RecommendOctober 5, 20262 min read

A dashboard answers the questions it was built for. The ones that come up in a meeting are usually just to the side of those: which of these moved this week and why, what is actually behind that score, is this a real signal or three articles repeating one source. The Recommend MCP connector puts those questions where you already ask them.

What it actually reaches

The connector exposes your demand data as tools an assistant can call: the projects in your workspace, the scope tabs inside them, every tracked object with its score and momentum, and the full detail behind any one of them. That detail is the part worth noticing. A score on its own is a number to argue about. A score with the question that produced it, the sources that supported it and the verdict on each one is a finding you can check.

How access works

You connect with OAuth, not an API key. You sign in to Recommend, choose which workspace the assistant may read, and approve the permissions on the consent screen. The assistant never sees your password, and the workspace you choose is the only one it can reach. Withdraw the authorisation from your account settings and it stops working immediately. It is read-only throughout: it can list and read, and it cannot create, change or delete anything.

What you can ask it

Three that come up most: which objects moved this week and what is driving them; show me the sources behind the top theme and say which of them actually support it; what does this scope look like against the same one a month ago. The assistant does the assembling. You get the answer with its evidence attached, in the place where you were already thinking about the question.In Claude, open Settings, then Connectors, then Add custom connector, and give it https://mcp.recommend.studio/mcp. Sign in, choose your workspace, approve. That is the whole setup. If your account also has a Predict Flow deployment, those tools appear alongside the demand ones; if it does not, they are simply not shown.

Connecting it

In Claude, open Settings, then Connectors, then Add custom connector, and give it https://mcp.recommend.studio/mcp. Sign in, choose your workspace, approve. That is the whole setup. If your account also has a Predict Flow deployment, those tools appear alongside the demand ones; if it does not, they are simply not shown.

If something does not work, or you want a workspace enabled for the connector, write to us at support@recommend.co.

Recommend for Claude: ask your demand data in plain language

What it actually reaches

How access works

What you can ask it

Connecting it

Related posts

AI in Marketing – Best Tools in 2026

How to Find Trending Topics (The Complete Guide)

How Platforms Like Recommend Are Reshaping the Speed and Quality of Digital Marketing

Turn your data into actions.