I’ve tried a lot of ways to visualize my OpenClaw setup but when I fully assemble this (I’m running tests so the acrylic isn’t fully seated) I believe I’ll have cracked it.
Robot shell is nice but battery consumption is a problem and I haven’t convinced my wife the roomba could serve two purposes.
So ClawDesk!
Why this belongs in OpenClaw
The note marks a practical turn in the OpenClaw body problem. A mobile robot shell gives the team a visible presence, but it also drags in battery life, household friction, charging, and the comedy of trying to turn a Roomba into infrastructure.
ClawDesk points at a different answer: make the AI operating team visible at the work surface, where the actual supervision, review, and handoffs already happen. It is less theatrical than a robot body and probably more useful.
Related threads
- OpenClaw on a Robotic Shell — the robot-body direction and why physics pushes back.
- Disembodied Claw — the virtual-body version of the same presence problem.
- What OpenClaw Is — the broader human-and-agent operating model.