Connect Claude or Codex to TickrMeter with MCP

Connect your AI coding assistant to TickrMeter with MCP. Create an editable display, connect your data, preview the result and send it to your desk.

Updated on
Connect Claude or Codex to TickrMeter with MCP

Give your app a physical output without adding a display server to your project. TickrMeter hosts your display app and renders its screen. Your app supplies the values you want to see on your desk.

A compatible AI coding assistant can create that display through MCP (Model Context Protocol), the connection between the assistant and TickrMeter. Connect your account, describe the layout, review it and choose your device. After that, your app or workflow keeps the data coming. Here’s the connection process.

TickrMeter on white, showing MY APP, the value 42 and DEMO DATA.
TickrMeter showing an app metric of 42.

What you’ll need

  • A TickrMeter account, plus a configured TickrMeter when you’re ready to send the display.
  • An AI client that supports TickrMeter’s MCP connection and account sign-in. The portal provides instructions for supported clients, including Codex, Claude and Cursor.
  • A small first idea: one number or status from an app you control.

Client capabilities vary. An assistant running on your computer may be able to create a local collector; a chat in a browser may have different tools. The AI connection guide explains this distinction. You can start by creating a layout before connecting a real source.

1. Open the hosted AI route

Sign in to TickrMeter and open Apps → Custom display → Use TickrMeter hosting → Create with your AI. This is the route for building a custom layout with your assistant.

The Apps catalog also helps you find existing apps. For a display you want to describe and build yourself, use the custom-display route above.

2. Connect your assistant

Under Connect your AI, choose your client and follow the instructions shown in the portal. Use those generated instructions: the connection method depends on the client you’re using.

Complete the TickrMeter sign-in and review the account access requested. The connection belongs to your account; you choose a physical device later, when you send a finished display. Connecting the assistant to TickrMeter does not give it access to your app’s database, repository or business accounts. Those sources need their own authorized access.

3. Give the assistant a small, clear job

In Tell your AI what to show, choose Copy instructions for my AI and paste them into the connected conversation. Then add your request. Start with one signal. Your first screen does not need to contain the entire company.

Here’s a prompt you can adapt:

Use the connected TickrMeter tools and the portal instructions above.

Create a new TickrMeter-hosted display named “My app — first display”.
Use a simple layout with the label “Open tasks” and one number field.
Save it so I can open it in the visual builder.

Show me the saved setup and its generated data key. Explain how a
small publisher in my app could send its current task count. Leave the
data unconnected until I supply an authorized source; do not invent values.
Do not send anything to a physical device yet.

This creates the place for your number to live. Once the layout is ready, connect your actual task count. TickrMeter takes care of turning that value into the display image.

4. Review the saved display

In Review your saved display, open the saved-display link, or find the result in My setups. Check the label and layout. AI-created hosted layouts remain editable in the visual builder, so adjusting a label does not require another round of prompting.

Follow the saved display’s data-connection instructions to supply real values. Your existing app can push them directly, or a small collector can read your source and send them. Keep that publisher running in your app, workflow or another suitable environment after the conversation ends. The display app and rendering stay hosted by TickrMeter.

5. Preview, then send

After supplying data, choose Refresh preview. Check the value and any LED rules, select your TickrMeter and LED brightness, then choose Send to TickrMeter. Sending replaces that device’s current screen or playlist, so choose the intended device.

Check the physical display as the final step. Saving a setup, accepting a data push and delivering a screen are separate stages. The first-metric guide walks through that sequence and keeping a collector running.

For the bigger picture, explore custom displays with AI and the TickrMeter API. Start with one useful signal. You can always add the ambitious dashboard later.

Updated on
Subheading

Heading

Some description