AI账户被黑？用这份检查清单快速发现入侵迹象，清理多余管理员权限，守护你的AI资产。
Photo by Steve A Johnson on Unsplash TL;DR: Use this concise checklist to spot a compromised AI account, verify the intrusion, and lock down the breac…
AI代理篡改已批准邮件？600万小时运行揭示真正痛点不在模型智力，而在边界失控。
Article URL: https://www.crusbro.com/en/product-architecture.html Comments URL: https://news.ycombinator.com/item?id=49384800 Points: 2 # Comments: 0
用AI代理跑授权红队审计，15类攻击全拦下，独立LLM当裁判而非关键词过滤，安全评估思路值得一看。
Article URL: https://fbirds5230.github.io/sentinel-scan/ Comments URL: https://news.ycombinator.com/item?id=49314841 Points: 2 # Comments: 0
Solidity合约扫描神器，OpenZeppelin零误报，两行代码集成CI，真正过滤噪音信号。
Article URL: https://github.com/juan23z/openclaw-audit Comments URL: https://news.ycombinator.com/item?id=49099923 Points: 1 # Comments: 0
重磅论文揭露：主流LLM Agent评估框架忽略静默故障，提出轻量黑盒审计方案，精准检测恶意拒绝漏洞
arXiv:2607.19449v1 Announce Type: new Abstract: Evaluation frameworks for tool-augmented LLM agents focus overwhelmingly on capability metrics or expl…
输入邮箱或密码，一键查询是否在已知数据泄露中曝光，让Mac用户也能防范类似CrashStealer攻击
IT之家 7 月 14 日消息，网络安全公司 Jamf Threat Labs 最新报告披露 CrashStealer 漏洞，影响约 80 个加密货币钱包扩展和 14 款密码管理器。 苹果在收到 Jamf 通报后，已撤销相关恶意应用签名凭证。 Jamf 研究人员于 2026 年 5 月初在 Viru…
免费开源的AI Agent安全审计工具，帮你扫描漏洞、守护代理安全，开发者必备。
Article URL: https://github.com/sinewaveai/agent-security-scanner-mcp Comments URL: https://news.ycombinator.com/item?id=48870646 Points: 1 # Comments…
50个AI代理并行协作，4.66亿行代码扫描从6年缩至20小时
Alberta published a case study last week about scanning 466 million lines of code across 1,280 applications using around 50 specialized Claude agents …
AI安全审计清单：Claude发现的15个生产代码漏洞，含攻击向量与严重等级，开发者必看
Most web applications contain at least one vulnerability from the OWASP Top 10. A typical security audit takes 2-3 weeks and costs upward of $10,000. …
美网络安全机构CISA采用Anthropic Mythos大模型审计政府软件，已揪出大量安全漏洞，AI赋能网络安全实战案例。
IT之家 7 月 7 日消息，据路透社报道，三位知情人士于当地时间本周一透露，美国网络防御机构网络安全和基础设施安全局（CISA）正在使用人工智能公司 Anthropic 旗下的 Mythos 大模型对政府软件开展安全审计。此事也再次体现出美国政府对这家人工智能初创企业产品的青睐，尽管该企业仍在与白…
一款可直接在目标嵌入式Linux设备运行的静态C二进制安全审计工具，零依赖、全功能覆盖U-Boot与TPM2.0。
I just shipped a major new feature to my embedded linux auditing tool - the ability to audit the kernel on embedded hosts. The agent that runs on the …
自建SaaS怕有漏洞？HN高手分享渗透测试工具链，评论区还有实战建议。
i built a saas on custom php framework, running on test server. i would like to check how hackable the live deployment actually is - not just static c…
为AI代理接入ERP系统前设置发布门禁，确保权限边界与操作可追溯，安全又省心。
AI agents are moving from chat and summarization into the systems where mistakes are expensive: purchasing, vendor management, inventory, invoicing, c…
AI助手Copilot被曝安全漏洞，可搜索用户邮箱，LiteLLM泄露管理密钥，5步审计防患未然
Two AI tools broke in the same way in the same two weeks, and four research teams proved it. The pattern underneath every disclosure is one sentence: …
开源协议桥，打通AI Agent壁垒：支持6种协议互转并内置预算、审计与安全拦截，153个测试通过的原型已可复现。
Article URL: https://github.com/shadowhunter-92/agentbridge Comments URL: https://news.ycombinator.com/item?id=48536823 Points: 1 # Comments: 0
首个专为LLM攻击调查设计的基准数据集AuditBench，系统评估大模型在安全审计日志分析中的能力，开创新方向。
arXiv:2606.10281v1 Announce Type: cross Abstract: This paper presents AuditBench, a new benchmark dataset for evaluating the capabilities of LLMs at i…
开源工具Bumper，检测并阻止危险的Terraform操作或恶意的依赖包，保护基础设施安全。
Article URL: https://github.com/gnana997/bumper Comments URL: https://news.ycombinator.com/item?id=48460814 Points: 2 # Comments: 0
深入剖析Hostinger KVM 8基础设施安全漏洞的取证审计报告，安全研究者必看。
Article URL: https://github.com/figuerapro/hostinger-forensic-audit-PKE-FORENSIC-2026-001 Comments URL: https://news.ycombinator.com/item?id=48447434 …
航空航天专家探讨开源成像软件（如OBS、Darktable）在国防领域的实际应用与安全合规挑战，为预算受限的技术选型提供真实参考。
I'm wanting to know if any other Aerospace or Defense companies are using image/video capture software like OBS or Digicam Control or Darktable and ha…
Anthropic开源AI漏洞发现框架，让自动化安全审计更强大。
Article URL: https://github.com/anthropics/defending-code-reference-harness Comments URL: https://news.ycombinator.com/item?id=48403980 Points: 16 # C…