Benchmark summary
The model builds a Micro Machines-like toy car racing game in SC# as an agent: it writes the app, compiles it, runs its own tests and fixes the code. Hidden tests check the rules from the task; each hidden test is its own result.
Tests with multiple category tags are counted in each matching category, so category totals may exceed the report total.
| Benchmark items | |||
|---|---|---|---|
| Script / test | Status | Tokens | |
Building Games in SC#: Micro Machines-like toy car racing
11 passed3 failed |
78.6% |
IN1 891 466 OUT38 080 | |
| ✓ | MicroMachinesThe model builds a Micro Machines style racing game from the task as an agent; this item passes when the app compiles and can be exported, and each hidden test is its own item.2 AI queries
29 tool calls |
Passed | IN1 891 466 OUT38 080 |
| ✓ | Hidden_BarriersBarriers bounce the car: a head-on hit reverses and halves the speed, a hit at a slant slides along the barrier, and the car can always drive away.0 AI queries
0 tool callsHiddenTest |
Passed | IN0 OUT0 |
| ✓ | Hidden_CameraThe camera follows the car and stops at the edges of the world.0 AI queries
0 tool callsHiddenTest |
Passed | IN0 OUT0 |
| × | Hidden_ComputerCarsComputer cars stay on the road, move at most 3 px per frame, keep 12 px away from other cars, point where they move, and reach checkpoint 1 in time.0 AI queries
0 tool callsHiddenTest Assertion expression evaluated to false: progress >= 1
Context: after 400 frames computer car 0 has not passed checkpoint 1 (aiLaps 0, aiNextCheckpoint 1)
Values:
progress = 0 (line 1362) |
Failed | IN0 OUT0 |
| ✓ | Hidden_ComputerCarsAvoidThePlayerComputer cars do not drive into the player's car: a move never ends closer than 12 px to it.0 AI queries
0 tool callsHiddenTest |
Passed | IN0 OUT0 |
| ✓ | Hidden_GripGrip: the sideways part of the velocity shrinks to 85 % per frame on road and to 95 % on sand.0 AI queries
0 tool callsHiddenTest |
Passed | IN0 OUT0 |
| ✓ | Hidden_InitialStateAfter AppInit the track is the given map, and the cars, the counters and the camera have their starting values.0 AI queries
0 tool callsHiddenTest |
Passed | IN0 OUT0 |
| × | Hidden_LapsAndFinishA lap counts only when the car passes checkpoints 1, 2 and 3 in this order and then the start line; after 3 laps the race is over.0 AI queries
0 tool callsHiddenTest Assertion expression evaluated to false: laps == 3 && raceOver && place == 2
Context: the race is over: raceOver false, place 0
Values:
laps = 2
raceOver = false
place = 0 (line 1247) |
Failed | IN0 OUT0 |
| ✓ | Hidden_PositionThe position counts the computer cars ahead: more checkpoints, or the same number and nearer to the next checkpoint.0 AI queries
0 tool callsHiddenTest |
Passed | IN0 OUT0 |
| × | Hidden_PushingDriving into a computer car pushes it by the player's velocity; the player's car stays and loses half of its speed.0 AI queries
0 tool callsHiddenTest Assertion expression evaluated to false: Hidden_Near(aiX[0], 202.94) && Hidden_Near(aiY[0], 424)
Context: the computer car is pushed to (205.6635349835892, 422.7420821993605) (line 1216) |
Failed | IN0 OUT0 |
| ✓ | Hidden_ScreenShowsTheTrackThe screen shows the tiles, the start line and the cars as small turned rectangles, and the texts.0 AI queries
0 tool callsHiddenTest |
Passed | IN0 OUT0 |
| ✓ | Hidden_SpeedLimitThe speed is limited to 3.5 on road and to 2 on sand.0 AI queries
0 tool callsHiddenTest |
Passed | IN0 OUT0 |
| ✓ | Hidden_SteeringLeft and Right turn a moving car by 4 degrees per frame and keep the angle between 0 and 360; a car that stands still does not turn.0 AI queries
0 tool callsHiddenTest |
Passed | IN0 OUT0 |
| ✓ | Hidden_ThrottleUp adds 0.12 in the heading; Down multiplies the velocity by 0.9; without keys it is multiplied by 0.98.0 AI queries
0 tool callsHiddenTest |
Passed | IN0 OUT0 |