遞迴遊戲創建器：一個代理式產品級體驗導向的遊戲框架

Recursive Game Creator: An Agentic Product-Level Experience-Oriented Game Harness

October 6, 2026

作者: Jiajun Chen, Haoyu Wu, Mingda Jia, Xihui Liu

cs.AI

摘要

近期遊戲設計代理在生成可玩遊戲方面已有長足進展。然而，程式正確性並不能確保玩家獲得愉悅的體驗。我們提出 Recursive Game Creator，一套以體驗為導向的框架，用以推動代理式遊戲開發從粗略的遊戲原型邁向具娛樂性的遊戲。Recursive Game Creator 圍繞四個元件組織遞迴式開發：Designer、Builder、Player 與 Reviewer。Designer 將使用者指令與 Reviewer 的回饋轉化為詳細計畫。Builder 將這些計畫轉化為候選遊戲。具程式設計原生能力的 Player 透過程式化介面建立並執行可重複使用的策略，以有效率地蒐集多樣化的遊戲遊玩軌跡，減輕因緩慢的圖形使用者介面式蒐集所造成的評估偏差。Reviewer 使用精心設計的軌跡式指標來歸納玩家偏好，並結合視覺證據與明確的文字偏好，依據遊戲特定準則評估遊戲。最後，Reviewer 接受較佳版本並提供改進審查以供下一回合使用，從而閉合遞迴迴圈。我們的方法在 GameCraft-Bench 上達到最先進的整體效能 77.89。在 GameASG-Bench 上，達到 53.2% 的嚴格任務成功率，相較於同模型基線提升 34.1%，且在所比較方法中具有最高的平均執行時檢查通過率 93.4%。一項使用者研究顯示更長的遊玩時間與更高的評分。程式碼即將發布。

English

Recent game design agents have made substantial progress in generating playable games. However, program correctness does not ensure an enjoyable experience for players. We present Recursive Game Creator, an experience-oriented harness to advance agentic game development from rough game prototypes into entertaining games. Recursive Game Creator organizes recursive development around four components: Designer, Builder, Player, and Reviewer. The Designer translates user instructions and Reviewer's feedback into detailed plans. The Builder turns these plans into candidate games. The coding-native Player creates and executes reusable policies through programmatic interfaces to efficiently collect diverse gameplay trajectories, mitigating evaluation bias caused by slow GUI-based collection. The Reviewer uses carefully designed trajectory-based metrics to induce player preferences, integrating with visual evidence and explicit textual preferences to evaluate games against game-specific criteria. Finally, the Reviewer accepts the better version and provides improvement reviews for the next round, closing the recursive loop. Our method achieves state-of-the-art overall performance of 77.89 on GameCraft-Bench. On GameASG-Bench, it achieves a strict task success rate of 53.2%, a 34.1% improvement over the same-model baseline, and the highest mean runtime-check pass rate at 93.4% among compared methods. A user study shows longer playtime and higher ratings. Code is coming soon.