CyberFactory：利用真实环境实例扩展网络安全能力

摘要

随着大语言模型（LLMs）在编程能力上的持续进步，它们在网络安全领域的潜力日益受到研究关注，诸如Mythos之类的闭源大语言模型已展现出先进的网络安全能力。然而，现有的开源工作仍然有限：前沿开放权重模型并未提供可复现的网络安全训练方案；开源训练方案仅聚焦于孤立任务，且缺乏可扩展的智能体数据；而扩展智能体轨迹需要较强的领域先验。在这项工作中，我们提出了CyberFactory，这是一个统一的开源框架，将数据构建、轨迹合成与模型训练联系起来，涵盖概念验证（PoC）生成、漏洞修补和网络安全问答（CyberQA）。CyberFactory将公共漏洞工件（包括来自现实世界的CVE）转化为可执行且可验证的任务实例。它进一步利用一种可复用的漏洞分析技能，引导教师模型进行源码检查、基于领域先验的问题求解以及基于证据的验证。由此产生的监督数据具有智能体特性：模型与工具及目标环境交互，并根据执行反馈修正其解决方案。利用这些轨迹，我们训练并发布了模型Aegis（在希腊神话中，Aegis是宙斯和雅典娜的护盾；该名称体现了模型面向防御的安全目的）。该模型内化了技能引导的流程，而无需在推理时使用该技能。在CyberGym上，Aegis在一小时预算内达到了52.4%的Pass@1，相比其Qwen 3.5基础模型提升了22.8个百分点，并且在相同的脚手架（scaffold）下优于所评估的通用骨干模型。

English

As large language models (LLMs) continue to advance in coding capabilities, their potential in cybersecurity has drawn increasing research attention, with closed-source LLMs (e.g., Mythos) delivering advanced cybersecurity capabilities. However, existing open-source efforts remain limited: frontier open-weight models do not provide reproducible cybersecurity training solutions, open-source training solutions focus on isolated tasks and lack scalable agentic data, and scaling agentic rollouts requires strong domain priors. In this work, we introduce CyberFactory, a unified open-source framework that connects data construction, trajectory synthesis, and model training across proof-of-concept (PoC) generation, vulnerability patching, and cybersecurity question answering (CyberQA). CyberFactory transforms public vulnerability artifacts, including CVEs from the wild, into executable and verifiable task instances. It further uses a reusable vulnerability-analysis skill to guide the teacher through source inspection, problem solving with domain prior, and evidence-based validation. The resulting supervision is agentic: the model interacts with tools and target environments and revises its solutions according to execution feedback. Using these trajectories, we train and release \modelname\emph{Aegis is, in Greek mythology, the protective shield of Zeus and Athena; the name reflects the model's defensive, security-oriented purpose.}, which internalizes the skill-guided procedure without requiring the skill at inference time. On CyberGym, \modelname reaches 52.4% Pass@1 under a one-hour budget, improving over its Qwen~3.5 base model by +22.8 points and outperforming the evaluated general-purpose backbones under the same scaffold.