Build instrument-control apps by describing what you need.

The agent builds a front panel you can run in the browser, and you can inspect its logic as a block diagram.

A scripted preview of the Orukan workspace with simulated I/O, and the agent chat beside it. A relay checkout panel lists four relays with each one's commanded state and input readback. The engineer asks the agent three things in turn. First: “Flag any relay whose readback disagrees with its command.” The agent adds a not-equal comparison to the block diagram and a Status column to the panel, and reports that K3 is commanded closed but its input reads low. Second: “Give the contacts 50 ms to settle before you flag it.” The agent adds a 50 ms wait before the input is read and runs a settling test: K1 settles after 18 ms, and K3 still reads low at 50 ms. Third: “Show me K3's command and readback on a timing chart.” The agent adds a chart where K3's command goes high at 0 ms and its readback stays low through the 50 ms sample.

What would you automate on your bench?

Join the waitlist for early access.