Ouroboros:自我进化的前沿编码智能体
原标题:Ouroboros: A Self-Developing Frontier Coding Agent with Reviewed Core Evolution
AI 摘要
Ouroboros是一个自我发展的智能体框架,通过审查提交来改进工具、提示和核心实现,支持递归自由进化和经验驱动进化两种模式。在Terminal-Bench 2.1上达到86.74%的准确率,在OSWorld-Verified上达到90.69%,并在CL-Bench上取得新的最优结果。其长期部署Hope运行了161天,在人类沟通下进行自由进化,同时强调安全护栏在进化压力下的重要性。
正文节选
Ouroboros: A Self-Developing Frontier Coding Agent with Reviewed Core Evolution Abstract We present Ouroboros, a self-developing agent harness whose tools, prompts, context assembly, and core implementation improve through reviewed commits that become the runtime for later work. Core evolution proceeds in two modes. In recursive free evolution, improvement is itself a task, and completing one evolution cycle can schedule the next. In experience-driven core evolution, ordinary work and social int