Skip to main content
Exec MCP connects your Exec workspace to the AI assistant you already work in. Ask a question and get an answer built from live Exec data. Any assistant that supports MCP can connect, and there is nothing to configure inside Exec. Once Exec is connected, your assistant can see your Exec information and act on it when you ask, the same way an app on your phone can use your camera or your contacts once you grant permission. It can hold several connections at once, so Exec sits alongside your calendar, your CRM, and your call recorder, and a single question can draw on all of them.

​
What you need before you start

Most people can set this up on their own in a couple of minutes. If your company is on a Team or Enterprise plan for your AI tool, an owner there adds Exec once before it appears for anyone else, and that is often a different person in a different department from whoever administers Exec.

​
Connect your AI tool

One URL, the same for every assistant, plus your normal Exec login.

Connect your AI tool

Setup for Claude, Claude Code, ChatGPT, Codex, Cursor, and any other MCP client

​
Try it

Once you are connected, ask this:
If your own recent sessions come back, you are connected and everything else on this page will work. If you get an empty answer or an error, see Troubleshooting.

​
Start where you sit

Three walkthroughs, each with a video. Pick the one that matches how you work.

Practicing on your own

Prep for the call you actually have, then check how you are doing.

Managing your team

Pull a readout, find who needs coaching, and build the practice that closes the gap.

Managing the org

Read the whole workspace, map skill strength by group, and summarize it for leadership.

​
What you can do with it

​
If you run enablement

  • Give me a pulse check on practice in my workspace over the last two weeks.
  • Which skill is my team weakest at, counting only skills where at least 20 people have been scored?
  • Which scenarios are people actually using, and which are being ignored?
  • Draft a short update for my leadership team on where our training program stands.

​
If you manage a team

  • How is my team doing? Give me each person's recent practice and their weakest skill.
  • Summarize Jordan's last five roleplay sessions and the feedback themes across them.
  • What is outstanding for my team, and draft a nudge for each person who is behind.
  • Build a roleplay based on the objection my team keeps losing to.

​
If you are practicing

  • How am I doing? Summarize my last five roleplay sessions.
  • What feedback keeps coming up for me?
  • Which skill should I practice next, and why?
  • Build me a roleplay to practice the thing I want to work on.
See the full set of prompts, grouped by role.

​
What the AI can see

It sees exactly what you see in Exec. Reps see their own practice. Managers see their groups. Admins see the workspace. Skill and scenario analytics are limited to admins and group admins, so a rep who asks for them is told no, even about their own scores. Their own sessions and feedback are still theirs to ask about. The only things anyone in the workspace can list are the member directory and the group list, the same as in the product, and nothing beyond that. Read the full detail, including what to tell your security team.

​
What it cannot do

A short list, and mostly reassuring.
  • It cannot change your data. No deleting, no editing, no assigning, no inviting people.
  • It cannot search your Knowledge Hub. It can use Knowledge Hub material as a source when it builds a roleplay, but it cannot read your Knowledge Hub back to you.
  • It cannot reach another company’s workspace. A connection is bound to one workspace.
  • It cannot do admin work. Anything you would do in workspace settings, you still do in workspace settings.
It can currently build roleplay scenarios, which is the only thing it writes back to Exec. You describe what you want, it drafts the scenario and shows you the draft, and nothing is published until you say so. Everything else it does is reading.

​
When to use MCP, and when to use Exec instead

Exec already has AI built in, so it is fair to ask why you would connect an outside assistant at all. Both are useful, for different things. Use Exec’s own AI when:
  • You are already inside Exec, reviewing a session or building a scenario
  • You want your Knowledge Hub searched, which MCP cannot do
  • You are new to this and want the guided version
Use MCP when:
  • You want Exec answers without leaving the tool you already work in
  • You are combining Exec with something else, like your calendar or your CRM
  • You want the assistant to do the writing too, such as drafting a nudge for each person who is behind
  • You are asking across many people or many weeks at once
A reasonable rule: start with Exec’s own AI, and reach for MCP when you are working at the scale of a team rather than a single session.

​
Admin controls

There is no MCP settings page in Exec today. People connect and disconnect from their own AI assistant, using their own Exec login and their own permissions. What that means in practice: you cannot currently see a list of who has connected, and you cannot revoke a connection on someone’s behalf from inside Exec. What you can rely on is that nobody sees anything through MCP that they could not already see by logging in, and that deactivating someone in Exec cuts their access on their next request. More on admin controls.

​
Troubleshooting

The three most common problems:
  • You are connected but every answer is empty. Usually the sign-in did not finish. Disconnect, reconnect, and watch for the Exec sign-in page to appear.
  • You asked about someone else’s sessions and got nothing. That is working correctly. See What the AI can see.
  • Charts do not appear. Charts need an assistant that can display interactive panels. In a text-only assistant you get the numbers instead.
Full troubleshooting guide.

​
FAQs

No. Anyone with an Exec account can connect. What you see through the assistant matches what you see when you log in, so a rep gets their own practice and a manager gets their team.
Most people can. On Claude Free, Pro, or Max, you add Exec yourself in a couple of minutes. Only on a Team or Enterprise plan does an owner have to add it once before anyone can connect. Nothing is needed from Exec’s side either way.
Not from inside Exec today. There is no MCP screen in the product, so there is no list of active connections and no way to disconnect someone else. If your company approves connectors centrally in your AI tool, that side is where connections are approved and can be turned off for everyone.
It can currently build roleplay scenarios, and it shows you the draft before anything is published. It cannot delete, edit, assign, invite, or change settings.
Yes. It refreshes quietly in the background and you should not have to sign in again for about a month of regular use. If you deactivate someone in Exec, or remove them from the workspace, their access stops on their next request rather than whenever the connection would otherwise have expired.
You pick one when you connect, and the connection is bound to it. To work in a different workspace, disconnect and reconnect, choosing the other one.