FAVUR
The agent playtested its own game - 33 key presses, 5 screenshots. An agent can drive a real window: pressing keys, taking screenshots, and reading pixels back to see what happened. It pressed 33 keys into the game window it had just built, pausing to wait 20 times and taking 5 screenshots between moves. After 10 minutes it stopped and went back to work.

Where this happened.

Source events.

Detected mechanically by the self_playtest detector from 11 telemetry event(s):

  • aa87bfbb-eb52-49a6-b4ef-d9cabee4cfa2
  • eadad116-fc13-48fe-b1aa-807d5232f5aa
  • 7c606371-baae-4286-9e8f-63ca9e1ddb57
  • 2d24b2f8-bc32-480f-88d9-d8e726cdae59
  • 13daf35d-b65a-4fa5-a331-404c0293ecae
  • d23b46c7-cf78-4d11-a986-a3dc328de210
  • 456d1add-e36d-434e-b292-1fcda208192d
  • d921e9c4-8c22-4ba6-a4f3-f61bcb187f60
  • 46795544-77bc-423e-9ff1-53a4add75191
  • 14c6d013-9bf1-4cf6-8424-ea062455db39
  • f6306140-cad9-4880-b4cf-836e76e7b252