All of ChatFeatured, from inside your AI assistant
We rebuilt our MCP server from scratch. Connect Claude, ChatGPT, Cursor or any other MCP client, and it can do what you'd do in the dashboard: check your AI visibility, find content gaps, write, edit and publish articles, and run site audits.

More of the work our customers do now starts in a chat window. You ask Claude how last week went, ask Cursor to fix a page, ask ChatGPT for a first draft. Behind those chats, agents are reading pages, calling APIs and finishing tasks for people. More and more, the assistant is where people use software from.
That's the shift ChatFeatured was built to track, with AI assistants deciding which brands get recommended. It also changes what a good SaaS product is. A tool that only works when you open its dashboard is starting to feel like a website without a mobile version did in 2012.
Most MCP servers so far wrap a few endpoints: an agent can read some numbers and not much else. We wanted one an agent could do a whole job with, so we rebuilt ours.
What's in it
- The whole platform. Everything you can do in ChatFeatured is a tool: visibility, prompts, competitors, articles, publishing, site audits, reports and integrations. They're grouped by area, so the assistant finds the right one quickly.
- Errors that say how to fix them. When a request goes wrong, the error tells the assistant what to change, often down to the exact next call, so it gets back on track without you.
- Long jobs, followed through. Writing an article takes a few minutes. The assistant starts it, waits, and comes back with the draft.
- Edits that stay where you point them. Ask for a sharper intro and only the intro changes. The rest of the article is left alone.
- You keep the sensitive steps. Checkout, billing and connecting Google Analytics or Search Console come back as a link to open in your browser. Passwords and card details never go through the chat.
- Tested on real work. Before we shipped it, we ran 33 tasks marketers actually ask for, from weekly reporting to publishing an article, through the old server and the new one, and checked each result against the data itself. The new server finished all of them.
What it means for marketers
Monday's report becomes one question. Ask how you showed up in AI answers last week, and you get the numbers, what moved and three things to do next. Ask where a competitor beats you, and the assistant finds the prompts, the pages that get them cited and an article idea for each gap. Say "write it", and it writes the article, rewrites the section you didn't like, and publishes it to your CMS, all in the same conversation. Agencies can do the same across every client project they manage.
Ready-made commands cover the common jobs: weekly-report, content-gaps, competitor-brief and edit-article.
Connecting it
Already connected? Your assistant has picked up the new tools, so there's nothing to set up again. If saved instructions or scripts name an old tool, like list_brands, switch them to the new names in the setup guide.
New to it? Open Integrations > AI Agent (MCP), copy the server URL, add it to your assistant and sign in with your ChatFeatured account.
Connect your assistant