Suitest MCP server

The Suitest MCP server lets an AI assistant work with your Suitest assets for you. You ask a question in plain language, and the assistant reads your apps, devices, app configurations, test packs and results to answer it.

What is it for?

MCP is short for Model Context Protocol. It is an open standard that lets an AI assistant call an external service in a structured way. The Suitest MCP server exposes your Suitest organization as a set of tools that the assistant can call.

Here are some things you can ask for:

  • Which devices in my organization are offline right now?
  • Which tests does the smoke test pack of my webOS app run, and on which devices?
  • Why did last night's test pack run fail on the Samsung TV?
  • Which test packs run on the Roku in our office, and when did they last fail?

The assistant works out which tools to call, calls them, and explains the answer. You never have to write the requests yourself.

How it works

Suitest hosts the server for you. There is nothing to install and nothing to keep running on your machine. Your client connects over HTTP to this address:

https://the.suite.st/api/public/v4/mcp

The server uses the same Personal or Service token as the Suitest Network API. Every call is scoped to the organization that owns the token, so the assistant can only reach the assets that token can reach.

Ready to set it up? See Connecting a client. For the full list of what the assistant can do, see the Tools reference.

MCP server or Network API?

Both talk to the same Suitest data. Pick the one that fits the job:

  • Use the MCP server when a person is in the loop. It is a good fit for exploring results, reviewing a test pack setup or debugging a failure with an assistant.
  • Use the Network API when a script or a CI job needs a stable, repeatable call.

The Suitest JavaScript API is a third option. Use it to program your test cases.

Good to know

A few things are worth keeping in mind.

The tools only read. The assistant can look at your data, but for now cannot create, change, delete or run anything (this will change in the future). To start a test pack or edit a configuration, for now, use the web app or the Network API.

You choose between live and pending. Edits to configurations and test packs wait in the app version change list until someone applies them. Ask the assistant to include pending changes when you want to review them. Otherwise it reads the state that is live right now. See Pending changes.

A test result comes in one piece. The assistant gets the line results, screenshot links, console log and network log of a test run all at once. See the Limitations.