cs.AI提交于 2026-10-07已译

Recursive Game Creator:面向体验的智能体产品级游戏开发框架

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

Jiajun Chen · Haoyu Wu · Mingda Jia · Xihui Liu

中文摘要

现有的游戏设计智能体在生成可玩游戏方面取得显著进展,但程序正确并不能保证玩家体验有趣。Recursive Game Creator 是一个面向体验的框架,将粗略的游戏原型推进为可玩的娱乐游戏。该框架围绕四个组件组织递归式开发——Designer、Player、Builder 与 Reviewer。Designer 把用户指令和 Reviewer 的反馈转译为详细规划;Builder 将规划实现为候选游戏;以代码原生方式工作的 Player 利用可编程接口创建并执行可复用的策略,高效收集多样化的游戏轨迹,缓解基于 GUI 的慢速收集所导致的评估偏差;Reviewer 采用精心设计的基于轨迹的指标来推导玩家偏好,结合视觉证据与显式文本偏好,按游戏专属标准评估游戏。最终 Reviewer 接受更优版本,并为下一轮提供改进反馈,闭合递归循环。该方法在 GameCraft-Bench 上取得 77.89 的整体最优性能;在 GameASG-Bench 上达到 53.2% 的严格任务成功率,较同模型基线提升 34.1%,并在对比方法中取得最高的运行期检查通过率 93.4%。用户研究表明其作品拥有更长的游玩时长与更高的评分。代码即将开源。

关键要点

01问题:现有游戏设计智能体虽能生成可玩游戏,但程序正确不保证玩家体验有趣,存在从原型到娱乐化产品的难题。
02方法:提出 Recursive Game Creator 框架,以 Designer、Builder、Player、Reviewer 四个组件递归迭代开发,Player 用编程接口高效收集轨迹并缓解 GUI 评估偏差。
03结果:在 GameCraft-Bench 取得 77.89 整体最优,GameASG-Bench 严格任务成功率 53.2%、运行期检查通过率 93.4%,均达 state-of-the-art。
04验证:用户研究显示玩家游玩时长更长、评分更高,佐证体验导向设计的实际价值。

解读

尚无解读。

原始英文摘要

arXiv:2610.08621v1 Announce Type: new Abstract: 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.