Panel 1 — A useful employee
Relay Roof, June 25, 2230, daytime. The receiver removed from the ration counter on June 21 remains in its tin. An unnamed ration clerk brings a stack of handwritten receipts. Tamsin keeps the device on the workbench.
CLERK: Can it go back? Hand receipts take time.
TAMSIN: First, we find out who it listens to.
Panel 2 — No public connection
Tamsin connects the preserved receiver to an isolated bench dial and a local test source. No tap, fee tray, public line or live account is attached. She sends one clock correction without a sender signature.
TAMSIN: Bench dial only. No tap. No bills.
CLERK: Who signed your test signal?
Panel 3 — An excellent applicant
The test dial moves back one minute. The clerk watches. This is a new bench experiment, not a replay presented as footage of the original incident.
TAMSIN: Nobody.
CLERK: It moved the minute anyway.
TEST RESULT: Unsigned correction accepted.
Panel 4 — References unavailable
The receiver goes back into its tin, disconnected. Tamsin records the single test; the clerk still has the handwritten receipts. No public reconnection is made.
CLERK: That was quick. Can we hire it?
TAMSIN: It only applies for management.
Closing observation
One receiver tested. Original sender still unknown.