All Use Cases

Your AI tools get better

Brilliant about the world, blank about your company

The AI tool you already use can reach your company records, the tools your team connected and the skills they saved. Nobody had to switch tools.

Powered By

The Challenge

  • Your AI knows everything about the world and nothing about your company
  • It cannot see the CRM, so you end up pasting the CRM into the chat
  • You become the integration: copying out of one system and into a chat window
  • Somebody two desks away already solved this, and your AI has no way to know
  • Whatever your AI can do stops at the edge of your own machine

The Runwork Solution

  • Your AI reads and updates your company records directly, instead of waiting for you to paste them in
  • It uses the tools your team already connected, on the connection somebody else set up once
  • It runs the workflows and scheduled work the team has built, when you ask for them by name
  • It saves a skill or a conversation to the workspace, so what you worked out does not stay in the window
  • When you ask for something the team already has, it is told to use the team's version rather than making a local copy

What your team gets

A question it can now answer

You ask about a customer and your AI reads the record, rather than asking you to paste it in first. The record is the same one your colleagues see.

The connection somebody else made

A teammate connected the CRM at the workspace months ago. Your AI uses that connection today, and you set nothing up.

A job you start by asking for it

The reconciliation the team built runs when you ask for it. You did not have to learn where it lives or what it is called underneath.

Work that leaves the chat window

You save what you just worked out as a skill. It stops being a thing that happened in your window and becomes a thing the team has.

A different ceiling

What your AI can do stops being what you can reach from your own machine, and becomes what your team has put in the workspace.

Where it still needs a terminal

Building and deploying a new app is command line work. What changes here is what your AI can reach and run, not what it can build for you.

A Day in the Life

9:00 AM. You connect it once. The AI tool you already had open gets pointed at your workspace over MCP. Nothing about how you work changes. What changes is the list of things it can reach.

10:00 AM. A question about a real customer. You ask what the last three orders were. Your AI reads the record rather than asking you to paste it, and the record it reads is the one your colleagues are looking at too.

11:30 AM. Somebody else's connection. You need something out of the CRM. A teammate connected it at the workspace months ago, so your AI uses that. You did not set anything up and you did not need a key.

2:00 PM. A job you ask for by name. The reconciliation the team built runs because you asked for it. You never learned where it lives.

4:00 PM. Something worth keeping. The long back and forth that finally worked gets saved as a skill. Your colleagues get it, and so do you, next month, when you have forgotten how you did it.

5:30 PM. The same tool, a different ceiling. You did not change AI tools today and you did not learn a new interface. What your AI can do is now bounded by what your team has built rather than by what is on your laptop.

Frequently Asked Questions

What can my AI actually do once it is connected?
Read and update your company records, use the integrations your team has connected, run the workflows and scheduled work they have built, reach the team's saved skills, and save a new skill or a conversation back to the workspace. In short: it can reach what your team has, and run what your team has already built.
Can my AI build and deploy an app for me?
No, and it is worth being exact rather than letting the rest of this page imply otherwise. Building and deploying a new app, schedule or workflow is command line work today. What your AI gains here is reach and the ability to run things that already exist, which is the part that works for somebody who never opens a terminal.
Do I have to tell it to use Runwork every time?
No. When your AI connects, it receives an instruction block telling it to use the team's shared version of something rather than making a local copy: the same guidance whether it reads it from a file on your machine or over the connection. So when you ask for something the team already has, it reaches for that.
Does this work in Claude Desktop and the ChatGPT chat tab?
Yes. Those are the surfaces where no instruction file exists on your machine, so the connection itself carries the context. That is the point of doing it over MCP rather than only through local files, and it is why this works for people who never installed anything beyond the AI tool they already use. 60+ AI tools are supported.
Does my AI see everything in the workspace?
It reaches what your workspace holds and what your role allows, in the same way you do. It is your access, used by the tool you are working in, rather than a separate set of permissions to keep track of.

Related Use Cases

Start keeping what your team figures out.

Start a 14-day free trial. No credit card required.