Benchmark summary
The model builds Klondike solitaire played with the mouse (drag and drop) with a polished look 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#: Klondike solitaire
14 passed2 failed |
87.5% |
IN1 214 030 OUT19 624 | |
| ✓ | SolitaireThe model builds a mouse-driven Klondike solitaire like the classic Windows 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.1 AI query
19 tool calls |
Passed | IN1 214 030 OUT19 624 |
| ✓ | Hidden_DealDeal() lays out the deck in the exact order: columns 0 to 6 with 1 to 7 cards, only top cards face up, the rest in the stock with deck[28] on top.0 AI queries
0 tool callsHiddenTest |
Passed | IN0 OUT0 |
| ✓ | Hidden_DragARunA press on the uncovered strip of a covered card drags it with all cards above it; a press on the top card drags only that card.0 AI queries
0 tool callsHiddenTest |
Passed | IN0 OUT0 |
| ✓ | Hidden_DragOntoAColumnDragging a card onto a card one rank higher of the other color moves it; the uncovered face-down card turns face up.0 AI queries
0 tool callsHiddenTest |
Passed | IN0 OUT0 |
| ✓ | Hidden_EnterStartsANewGameEnter starts a new game at any time; a mouse press in that frame does nothing, and a drag in progress is dropped.0 AI queries
0 tool callsHiddenTest |
Passed | IN0 OUT0 |
| ✓ | Hidden_FoundationCardBackThe top card of a foundation can be dragged back onto a column when the column rule allows it.0 AI queries
0 tool callsHiddenTest |
Passed | IN0 OUT0 |
| ✓ | Hidden_FoundationsFoundations take one card of their own suit in rank order from the Ace; other drops go back.0 AI queries
0 tool callsHiddenTest |
Passed | IN0 OUT0 |
| ✓ | Hidden_IllegalDropsGoBackIllegal drops and drops outside every pile put the cards back and are not moves; only the mouse point decides the target.0 AI queries
0 tool callsHiddenTest |
Passed | IN0 OUT0 |
| ✓ | Hidden_KingOnAnEmptyColumnOnly a King, alone or with the cards above it, can be dropped on an empty column.0 AI queries
0 tool callsHiddenTest |
Passed | IN0 OUT0 |
| ✓ | Hidden_NewGameDealsAllCardsA new game (AppInit) deals all 52 different cards: 28 in the columns as Deal() does and 24 in the stock.0 AI queries
0 tool callsHiddenTest |
Passed | IN0 OUT0 |
| ✓ | Hidden_RightClickA right click sends the waste top card or a column's top card to its foundation when legal; otherwise nothing happens.0 AI queries
0 tool callsHiddenTest |
Passed | IN0 OUT0 |
| × | Hidden_ScreenDragFollowsTheMouseWhile dragging, the dragged card is drawn under the mouse at the pressed offset; after the drop the table looks as before.0 AI queries
0 tool callsHiddenTest Assertion expression evaluated to false: Hidden_Differ(away, back) >= 64 * 90 * 5 / 10
Context: while dragged the card is not drawn in its column (line 1521) |
Failed | IN0 OUT0 |
| ✓ | Hidden_ScreenShowsCardsFace-up cards, face-down cards and the empty table look different.0 AI queries
0 tool callsHiddenTest |
Passed | IN0 OUT0 |
| ✓ | Hidden_StockAndWasteA click on the stock turns its top card to the waste; on the empty stock it turns the whole waste back in order; each is a move.0 AI queries
0 tool callsHiddenTest |
Passed | IN0 OUT0 |
| × | Hidden_StripEdgesA press picks the card whose uncovered strip contains the mouse, exactly at the strip edges.0 AI queries
0 tool callsHiddenTest Assertion expression evaluated to false: tableau[3].Count == 5 && Hidden_Same(tableau[1], [Hidden_Card(4, 0), Hidden_Card(6, 2), Hidden_Card(13, 3)]) && hiddenCount[1] == 2
Context: column 1: 3 31 51
Values:
tableau[3].Count = 5
hiddenCount[1] = 1 (line 1326) |
Failed | IN0 OUT0 |
| ✓ | Hidden_WinningThe last card on the foundations wins: won is true, "YOU WIN" is shown and the mouse does nothing more.0 AI queries
0 tool callsHiddenTest |
Passed | IN0 OUT0 |