Mozilla 用 Claude Mythos 在 Firefox 代码库中修复 400+ 安全漏洞
https://t.co/7YnqMnFefK
Mozilla 工程师亲述怎么让 Claude 在千万行代码里挖出 400 多个安全漏洞,包括十多年没被发现的老 bug。重点是他那套可复制的 harness:goal/loop 循环、验证器过滤误报。自己仓库里有安全债的工程师真该看看。
Mozilla 在 Claude Mythos 模型发布前(该版本随后被撤回)就已将其用于 Firefox 约 1000 万行代码的测试。测试结果是修复了超过 400 个安全漏洞,其中一些漏洞已经在代码库中潜伏了十多年。Mozilla 杰出工程师 @bgrins 在访谈中拆解了这套智能体化漏洞挖掘系统。
技术上,这套系统并非简单把模型丢进代码库:Mozilla 采用了目标/循环(goal/loop)模式来驱动智能体,用独立的验证器(verifier)过滤误报,并通过"对智能体说谎"等提示策略引导模型行为。Brian 认为最终效果一半来自 Mythos 模型本身,一半来自这套系统搭建。
Brian 强调这套 harness 并不神秘,开发者用一个下午就能写出类似的系统。这对想在自有大型代码库上跑漏洞扫描的工程团队是一个可直接复制的参考流程。
https://t.co/7YnqMnFefK
x.com/clairevo/statu… claire vo 🖤 @clairevo Before Fable got released (and pulled) @mozilla was quietly testing Claude Mythos against Firefox's 10M line codebase. The result? Over 400 security bugs fixes, including ones that had been hiding in the codebase for over a decade. @bgrins , distinguished engineer at Mozilla, walked me through the agentic bug-finding harness behind the model. His take? It was 50% mythos / 50% setup. In this ep, Brian walks through: - why you can't just point a model at 10M lines of code - how to write a good goal/loop pattern - killing false positives with a verifier - why it's good to "lie to the agent" And guess what? This isn't magic - you can write your own similar harness in less than an afternoon. Watch now on YT: youtu.be/Idjt53tTv2U
This report is categorized under AITOP's "How-to" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include Claude, Mozilla, Firefox, agents.
Source attribution: curated from Claude (@claudeai).