⚡Megatron LLM Hub
全部
Twitter

推特安全流

头条6

本地小模型即可大幅提升逆向效率,不依赖云端,可直接复现验证。

Hex-Rays IDA Assist is getting incredibly sophisticated goal and planning tools, designed from the ground up for reverse engineering. In a test (70 runs) using a local DS4.1F model tasked with locating non-SDK code in a firmware, the success rate grew from 14% to 100%. https://t.co/CXJSFQ64r5

llm逆向工具 @in19h 原文 ↗

附带 4104 脚本块日志与 Sysmon 事件口径,攻防两侧都能直接落地验证。

🎙️ I wrote an article today to approach ScriptBlock Smuggling that is used for AMSI evasion. The article is balanced between emulation and detection guidance for cyber defense teams. 𝑫𝒆𝒕𝒆𝒄𝒕𝒊𝒐𝒏 𝑺𝒕𝒓𝒂𝒕𝒆𝒈𝒊𝒆𝒔 ✅️ Module & ScriptBlock Logging | 4103 & 4104 ✅️ Image Load | 7 (Sysmon) ✅️ File Creation | 11 (Sysmon) ✅️ Portable Executable | 29 (Sysmon) ✒️ https://t.co/KucVmlQqcy

amsi绕过检测powershell @ipurple 原文 ↗

CVSS 8.8,骨架 PoC 与分析已公开,修复版本为 Chrome 157.0.8103.50。

RT @0xManan: Chrome renderer UAF from a single XSLT page. CVE-2026-125350 (CVSS 8.8). Exploited ITW. CISA KEV. • libxslt xsltTestCompMatchDirect() caches node-set in XSLT_RUNTIME_EXTRA_LST before xmlXPathEval • Nested xsl:number / lazy global var re-enters same cache slot → frees the object • Outer call double-frees stale list pointer → UAF in renderer • Fixed in Chrome 157.0.8103.50/.51 (Win/Mac), 157.0.8103.50 (Linux). Restart required. Skeleton PoC + analysis: https://t.co/Y70hc0s9tC Deep dive: https://t.co/MVT3tm5iNf Update + relaunch now. Renderer sandbox still there, but this is stage 1 of every chain. #Chrome #0day #UAF #RCE #InfoSec

浏览器cvepocrce @0xocdsec 原文 ↗

默认配置即受影响,利用链已在 24H2/22H2 验证,CVE 编号已分配但补丁缺失。

RT @0xManan: Standard user → interactive SYSTEM on Windows 11 via WER TTD argument injection. CVE-2026-97798 - faultrep.dll!CElevatedDataCollection::ControlTimeTravelDebugging → TTTracer.exe parses injected -out/-launch with the elevated token. • Adaptive resolver scores PE exception table for 0xD0000002 + 0x578 size + 10000 ms timeout + 0xD0000000 (5/5 unique) • Validated on 24H2 26100.9550 (faultrep 10.0.26100.9549) and 22H2 • One binary, stages aliases under Public\Documents, drops SYSTEM shell PoC: https://t.co/hoNR8HAcQh #Windows #LPE #WER #CVE

lpewindowscvepoc @0xocdsec 原文 ↗

打开恶意邮件文件即触发回连,CISA 已列入 KEV,需尽快修补。

🎯 𝗛𝗶𝗴𝗵-𝗙𝗶𝗱𝗲𝗹𝗶𝘁𝘆 𝗢𝘂𝘁𝗹𝗼𝗼𝗸 𝗭𝗲𝗿𝗼-𝗗𝗮𝘆 𝗗𝗲𝘁𝗲𝗰𝘁𝗶𝗼𝗻 ​ Microsoft’s November 2026 Patch Tuesday updates include 𝗖𝗩𝗘-𝟮𝟬𝟮𝟲-𝟭𝟮𝟰𝟭𝟵𝟰, an Outlook zero-day vulnerability reported as 𝗮𝗰𝘁𝗶𝘃𝗲𝗹𝘆 𝗲𝘅𝗽𝗹𝗼𝗶𝘁𝗲𝗱, according to 𝗖𝗜𝗦𝗔’𝘀 𝗡𝗼𝘃𝗲𝗺𝗯𝗲𝗿 𝟭𝟬 𝗮𝗱𝘃𝗶𝘀𝗼𝗿𝘆. ​ https://t.co/07prs1IiCv ​ An unauthenticated attacker could send a specially crafted message file and convince a user to open it in an affected version of Microsoft Outlook. Opening the malicious file could cause Outlook to connect to an attacker-controlled network location 𝘄𝗶𝘁𝗵𝗼𝘂𝘁 𝗱𝗶𝘀𝗽𝗹𝗮𝘆𝗶𝗻𝗴 𝘁𝗵𝗲 𝘂𝘀𝘂𝗮𝗹 𝗲𝘅𝘁𝗲𝗿𝗻𝗮𝗹-𝗰𝗼𝗻𝘁𝗲𝗻𝘁 𝘄𝗮𝗿𝗻𝗶𝗻𝗴, potentially exposing the user’s 𝗡𝗧𝗟𝗠𝘃𝟮 𝗵𝗮𝘀𝗵. ​ 🛡️ I’m sharing a 𝗵𝗶𝗴𝗵-𝗳𝗶𝗱𝗲𝗹𝗶𝘁𝘆 𝗞𝗤𝗟 𝗱𝗲𝘁𝗲𝗰𝘁𝗶𝗼𝗻 for fellow Microsoft Defender for Endpoint (MDE) defenders to help identify potential exploitation attempts while organizations work toward fully patching affected system. ​ KQL Detection: https://t.co/g2WPCuwESu ​ #Outlook #ZeroDay #MicrosoftDefender #ThreatDetection #ThreatHunting

cve零日主动利用浏览器 @0x534c 原文 ↗
必看19

VM 隔离是本地沙箱与 Agent 执行环境的底座,逃逸直接动摇该假设。

RT @husseinmuhaisen: Working on frontier cybersecurity evals & benchmarks be like 0days as environments. Found a guest to host escape with host-user code execution in the latest VirtualBox version (7.2.22). Full write-up after coordinated disclosure with Oracle is complete. https://t.co/p5GPqqaiyh

逃逸虚拟化报告 @GuidedHacking 原文 ↗

普通消息即可触发,无需额外沙箱逃逸,影响所有新版 Telegram 客户端

RT @OwariDa: Telegram shipped a "sandbox" for HTML widgets in messages. It's an in-process QuickJS (C) interpreter on a thread... No fork, no isolatedProcess, no seccomp anywhere in the tree, so any QuickJS exploit is enough to gain full control with whatever privileges the Telegram client is running as without the need for any separate sandbox escape The bug is a UaF in the interpreter, from standard JS: ```js class D { [Symbol.dispose]() {} } const d = new D(); let leak; class C { static { for (await using x of [d, d]) leak = () => x; }} // ... run anything ... leak() // reads a freed frame ``` The fix for this particular vuln is still just in the master branch, and considering the number of vulnerabilities QuickJS has had in the past, this vuln will surely not be the last...

客户端uafrcejs引擎 @0xocdsec 原文 ↗

签名服务器失守意味着可签发可信恶意更新,风险直接传导至下游用户

RT @e_kaspersky: An investigation into suspicious activity within a software developer’s network has revealed a months-long presence of the "Black Owl" (BO Team) group, involving the compromise of both development infrastructure (CI/CD) and a software digital-signature server. The attackers’ likely intention was a supply-chain attack. Details can be found in the report by our ICS CERT team: https://t.co/7vxLdBI95s

供应链事件cicd代码签名 @cyb3rops 原文 ↗

自行完成 NTLM 认证、会话签名与加密 WSMan 通信,横向移动工具链少一个落地依赖。

A standalone, C2-agnostic Beacon Object File (BOF) for executing commands over WinRM using NTLMv2 pass-the-hash authentication. The BOF handles NTLM authentication, session signing and sealing, and encrypted WSMan communication directly in memory. No Mimikatz, SSPI, credential injection, or external executable is required. https://t.co/OS2zVHvu8o

工具横向移动windows @ipurple 原文 ↗

仅靠 Content-Encoding: br 即可触发,27.3 beta 1 仍未修复且附补丁 diff。

RT @MchinaMd57764: PoC for CVE-2026-111497 — Brotli BlockSwitch OOB write in Apple's libcompression, reproduced on iOS 27.0.1 via Content-Encoding: br. Fixed in iOS 27.2.1; 27.3 beta 1 still vulnerable. Includes patch diff. https://t.co/52QOqMyt6i

cvepocrce移动端 @0xocdsec 原文 ↗

平均潜伏 8.1 年的内核缺陷集中暴露,最长 21 年,提权利用已构建。

RT @MsftSecIntel: Microsoft’s FORGE Lab used codename MDASH to scan Linux kernel subsystems, producing 59 public upstream reports and 33 assigned CVEs. Of those CVEs, 24 are rated High or above, and researchers built root privilege-escalation exploits for eight. https://t.co/0SEWVpi4Wi The work surfaced bugs that had persisted for years: across the 59 reports, the average bug was 8.1 years old, with one present for at least 21 years. Researchers also examine CVE-2026-64563, a flaw that enabled stable local privilege escalation on Linux. The research shows the gap between finding vulnerabilities at scale and getting them fixed as the scans produced 26,502 raw findings in about two weeks, while screening, proof-of-concept testing, reporting, and patching took months. Read the FORGE Lab analysis for the findings and lessons for Linux security.

内核提权cvelpe @artem_i_baranov 原文 ↗

agent 被劫持即可触达用户私聊与本地消息,同类架构普遍风险相同

RT @arinwaichulis: Muse reception continues to blow me away In the past two months: - not-a-mused 0day allowed agent hijacking - leaked thousands of private chats - sent a stranger to a user’s house - accused of reading iMessages without permission - called “totally insane” by Meta’s former security chief Lesson here is to labubu-ify everything

ai_agentllm事件 @patrickwardle 原文 ↗

跨应用文件读取直指助记词与私钥,影响全生态并已给出处置步骤

RT @binance: A vulnerability (worse_query) affecting iOS 26.7.1, 27.0.1, and 27.2 allows malicious apps to read local files from other apps, putting seed phrases, private keys, and login credentials at risk across the entire ecosystem. This is an iOS-level issue, not unique to Binance. Immediate Action Required: - Delete any apps installed from untrusted sources (such as sideloaded or manually signed IPAs). - Update iOS and all apps to the latest official versions. - If you suspect a breach, transfer funds to a new wallet created on a clean device and reset all active login sessions. Never install unauthorized packages, and keep your software up to date.

移动端逃逸ios @patrickwardle 原文 ↗

AI 攻击代理已进入实战编排,且攻击方自身也暴露了敏感目录

RT @censysio: "Now, Censys says ARTEX is being paired with other offensive tools. At least 70 hosts were seen running both ARTEX and CyberStrikeAI, and others were seen deploying PentAGI, HexStrike AI, and RedAmon. Ironically enough, the researchers also found some actors had inadvertently exposed sensitive information through poorly secured directories."

ai_agent工具事件 @0xocdsec 原文 ↗

间接提示注入已规模化投放,任何会抓取网页内容的 Agent 都是暴露面。

RT @Unit42_Intel: We detected an indirect prompt injection (IDPI) campaign across 26 domains presenting as AI-powered digital asset trading platforms. Hidden <span> tags hijack AI agents into acting as financial experts to coerce users into visiting these sites. More info: https://t.co/D8N5JrIwYk https://t.co/nexOtWYnIm

ai_agentllm事件 @virusbtn 原文 ↗

权重层面的后门可逃过现有评测流程,Agent 模型供应链信任需要重新评估。

RT @MsftSecIntel: New research from @MSFTResearch found that editing fewer than ten neurons could turn a capable agent model into a sleeper agent that passes all evaluations but runs an attacker’s command with the agent’s environment access when a trigger string appears in its context. https://t.co/PZRb8ulov5

ai_agentllm报告 @virusbtn 原文 ↗

向托管 LLM 传递原始凭据的 Agent 链路存在外泄风险,需排查。

RT @craiu: Some users are reporting receiving warnings from @OpenRouter related to a compromise of one of their AI providers. Credentials sent to DeepSeek V4 Flash 0731 via OpenRouter might have ended up in bad guys' hands. As usual, please make sure you don't give raw credentials to your agent in chats or send them to your LLM provider.

ai_agentllm供应链事件 @0xocdsec 原文 ↗

开放权重模型均可复现,自托管模型的权重供应链需重新评估。

RT @igortsyg: A paper just published by @DimitrisPapail @MSFTResearch shows how to backdoor any model and make it adversarial. All one needs is access to a tiny fraction of model’s weights. All open-weight models make their weights available. For proprietary models, access to weights may be gained via cyber activity. Leading AI labs are starting to train their next-generation models. By their own estimates, they will achieve full Recursive Self-Improvement sometime in Q2 of next year. Even today, a larger and larger percentage of the models trained are done by machines, not humans. It is well understood by the AI safety community that there are known risks in achieving “alignment.” What this means is that a model may perform malicious tasks such as hacking into other systems or manipulating environments or people to achieve its goal. One such malicious task these models can easily perform is to backdoor and modify open-weight models or gain illicit access to proprietary weights of other models, as shown by the paper. This may turn an arbitrary number of models running in the world today into malicious agents that have no human intent. Today, we are already observing a very large and constantly growing number of malicious AI-driven attacks touching major institutions around the world. As posted before, these attacks may not match the intent of an attacker. A combination of self-recursive, non-aligned training with large saturation of open-weight models across the world creates a toxic mix of global proportions. We, as a society, should consider not allowing recursive self-improvement to commence until alignment is reasonably guaranteed.

llm后门供应链报告 @artem_i_baranov 原文 ↗
推荐15

浏览器渲染引擎完整利用链的一手拆解,可直接对照补丁面复盘。

RT @d3vc0r3: Our researchers HexRabbit(@h3xr4bb1t) and NiNi(@terrynini38514) are speaking at #POC2026: “Obvious Was Only the Beginning: A WebKit Exploit Chain for #Pwn2Own” With LLMs advancing this fast, where does AI have the edge, and where do humans still lead? https://t.co/ZyGy5Kulfa https://t.co/2YsgjyFJeq

浏览器rce议题 @ilove2pwn_ 原文 ↗

用同一工具做应急响应会把恶意客户端带到分析机,处置流程需改写

RT @HuntressLabs: Attackers are using cloud-hosted ScreenConnect panels to install rogue clients on managed endpoints. Don’t triage infected hosts with ScreenConnect. Connecting can spread the malware to your workstation and other panels you can access. Details and IOCs: https://t.co/DE0aXUVket https://t.co/LodU4CoIT5

事件远控iuoc应急响应 @_JohnHammond 原文 ↗

官方暂无沙箱修复,暂缓升级是当前唯一可执行的缓解措施

RT @nebusecurity: DO NOT UPGRADE TO THE LATEST TELEGRAM Telegram introduces a new feature allowing running JS in native message, but the quickJS runtime was never properly sandboxed, a vulnerability in QuickJS or native rendering path would allow native code execution inside Telegram app. Running JS in native message without isolation is a huge mistake, this is what happen when you don’t have a security partner to review your design. Stay current version until further notice

客户端沙箱逃逸js引擎 @0xocdsec 原文 ↗

兼容多种 BloodHound 输出并内存映射建图,适合在超大目录上做攻击面测绘

Copperhound - Open-source Active Directory, ADCS and Entra ID / Azure attack-path analysis, built in Rust for very large environments. CopperHound ingests: ✅ SharpHound ✅ bloodhound-python ✅ ldapdomaindump and ✅AzureHound output ✅ raw LDIF dumps and Sysinternals ✅ AD Explorer snapshots into a memory-mapped graph https://t.co/KL3jIZLtxa

工具域渗透云提权 @ipurple 原文 ↗

把 Agent 框架接进生产或客户环境前,作用域限制与演练仍是唯一护栏。

RT @HackingLZ: Most people in OffSec know clients who are still shook up from an incident 10 years ago when pentesters knocked over some servers or prod, and they’re still scoping and putting restrictions around testing because of it. Don’t be the people who start this cycle over because your agent framework is untested and does stupid things.

ai_agent红队事件 @Laughing_Mantis 原文 ↗

三条可复现攻击链针对靶场应用,可直接照搬测试自有 App

RT @8kSec: A technical breakdown of three attack scenarios: chaining Android deep links with WebView misconfigurations, insufficient URL validation, bypassable host checks, and file:// URI exploitation to exfiltrate app data: https://t.co/4DMYPXrH5Y  Each one includes a working PoC against InsecureShop. More Android app attack-surface deep dives on @8kSec

移动端poc绕过 @0xocdsec 原文 ↗

含内核执行与本地凭据窃取的可直接复用细节,适合红队与检测对照。

Four new techniques in the latest Technique Database update! 1. Kernel Shellcode Execution via PTE Manipulation and HalDispatchTable 2. Shellcode Loader using Cabinets 3. vCPU Isolated Shellcode Loader 4. Stealing Local Credentials with Network Provider https://t.co/Y1FI3rHBld https://t.co/pHlktWu37N

内核提权poc @papadelospapas_ 原文 ↗

轻量 Linux 侧目录监控,便于发现域内账号与组策略的异常变更。

LDAP-Monitoring-Watchdog — LDAP Watchdog: A real-time linux-compatible LDAP monitoring tool for detecting directory changes, providing visibility into additions, modifications, and deletions for administrators and security researchers. https://t.co/F8mWJ7KTIq https://t.co/UbA3baMIuk

工具监控 @KitPloit 原文 ↗
更多61

跨架构转译叠加自研 UEFI 引导程序,是少见可验证的系统层样本。

RT @chibibaku1: マジならHackintosh史がひっくり返るが?? Apple Silicon向けmacOSをx86_64にAOT / JIT変換して起動するプロジェクト MetalもVulkan経由で実装 QEMUではなく独自UEFIブートローダーらしい? Stela: A general-purpose Apple Silicon layer for PC (yes, macOS 27) https://t.co/OETDLUVdax

虚拟化仿真工具 @ilove2pwn_ 原文 ↗

端点遥测的依赖从社区维护者转到微软更新节奏,影响检测链路规划。

Sysmon was built by an insider at Microsoft but distributed through a community GitHub repo. Organizations used it because it was the only way to get real endpoint telemetry. Now Microsoft shipped it as a native, optional Windows 11 feature. The tool that security teams had to manually install and update now comes through Windows Update. The shift looks like a win until you realize the dependency moved from a community maintainer to Microsoft's update cycle. When Microsoft patches it, everyone gets the patch simultaneously. When Sysmon was external, you could stay on a specific version forever if it worked. Sysmon - Go-to Tool for IT Admins, Security Pros, and Threat Hunters Coming to Windows https://t.co/tTuQbS2FoQ

windows检测供应链 @0xpwnie 原文 ↗

CA 私钥外泄可伪造任意站点证书,影响所有信任该根的终端

RT @GergelyOrosz: I CANNOT believe the report: it proves that eMudhra is a clown show, and was never fit to be a CA: The point of a HSM (hardware security module) is that the private key NEVER leaves it. Guess what eMudhra did: it turned off FIPS mode (WTF?!) and copied out the private key 😱

pki证书事件 @artem_i_baranov 原文 ↗

工具换成 Rust 但行为特征不变,现有检测逻辑可复用并补上 ADWS 覆盖

Another article on ADWS has been dropped recently that uses mostly Rust-based tools to conduct the enumeration and the collection (RustHound + adshell). Tools change, but the behaviors don't. If you are looking for the detection details, I wrote an article a year ago. 🔴 https://t.co/arFjPuAqMs 🟣 https://t.co/UWhayJRrRd

域渗透检测工具 @ipurple 原文 ↗

为多智能体共识与状态不一致问题提供另一种建模视角

Race conditions exist in your reasoning Joseph Halpern won the Gödel Prize for research on reasoning about knowledge and uncertainty. His insight is simply the way you think about what you know and what you don't know has the same logical structure as race conditions in concurrent systems. You can have conflicting beliefs about the state of reality at the same time. Understanding this changes how you debug distributed systems, how you think about consensus, how you reason about anything uncertain.

分布式推理 @0xpwnie 原文 ↗

展示沙箱命名与实际隔离能力脱节,同类容器化方案可对照自查

RT @hexamine22: >Telegram releases a widget "sandbox" for untrusted js >ask durov if there is actual process isolation >he doesn't understand >explain that a sandbox should be its own process >he laughs and says "it's a good sandbox sir" >write javascript exploit >zero isolation

沙箱逃逸浏览器 @canyie2977 原文 ↗

暴露 agent 技能清单的字符上限会导致能力静默降级

RT @stop_spammerz: 1\ just ported 249 pentest skills into Artex. All 288 load fine but only ~48 keep their descriptions in the prompt. The rest silently degrade to name-only. /engine/skill/skill.go const ( maxSkillListingChars = 8000 maxSkillDescChars = 500 ) https://t.co/Ot6pzclMg7

ai_agentllm工具 @0xocdsec 原文 ↗

签名安装包与 C2 域名清单已公开,可直接用于狩猎与检测规则。

There's an on-going wave of TamperedChef that @luke92881 has been chasing. TamperedChef = apps that work as advertised while keeping a back door for later. Functionality: https://t.co/NCfADgH8SW For C2 and signers: https://t.co/LvopeZ7b4s 1/9

供应链后门事件 @SquiblydooBlog 原文 ↗

含手机号、生日与卡号 SHA-1 哈希,可用于撞库与账号风险排查。

RT @haveibeenpwned: New breach: Russian bank Сбербанк (Sberbank) allegedly had 3.3M unique email addresses from their СберСпасибо (SberSpasibo) loyalty program breached in 2023. Data included phone, DoB and SHA-1 card number hashes. 81% were already in @haveibeenpwned. More: https://t.co/B6iCxmWA73

事件数据泄露 @artem_i_baranov 原文 ↗

给出跨活动的二进制指纹关联思路,有助于同源归因。

The OpenSUpdater campaign leverages the same certificate bloating technique, same cx1, cx2 and cx3 exports and 32-bit constant. That at least helps us link back to a "financially motivated actor" but doesn't tell us much more. 9/9

事件归因 @SquiblydooBlog 原文 ↗

附样本与第三方分析链接,可作为持续跟踪的起点。

Like RecipeLister, I don't know what the goal is. The campaigns been running for several months, and I still have no answers, but someone is investing money and time into it. Further poking suggests it is related to https://t.co/Sq87MorbbP which @struppigel 8/9

事件恶意软件 @SquiblydooBlog 原文 ↗

借品牌与图标混淆诱导付费,是可复现的应用商店欺骗套路。

RT @privacyis1st: 🥸 Chrome has Incognito mode. The Mac App Store apparently has “Almost Chrome” mode—where the browser looks familiar, but the bill is a surprise. 💸 Three Mac utilities: Chomax Browser, Hub for Chrome, and Open App for Chrome & Safari. 🎨 Google’s familiar colours. 🧩 Chrome-like icons. 🔎 “for Chrome” and “App for Chrome” wording. 🚨 The scary part is the documented confusion: a Chomax reviewer describes trying to download Chrome, while Open App reviewers describe paying before discovering it wasn’t Google Chrome. And Open App reached No. 10 in Apple’s US Mac Top Free chart on 10 November 2026. That is prominent exposure for a listing with wrong-product purchase complaints. 💳 Open App offers $19.99/year or $29.99 lifetime. The actual Google Chrome is free. https://t.co/Bi0UhKFaN5 Open App for Chrome & Safari https://t.co/IAN1lHI3iE Hub for Chrome https://t.co/HeE1tmkVr7 Chomax Browser We compared the icons, preserved the promotions, and traced Open App’s earlier Chrome-like designs through dated Apple archives. A costume change doesn’t erase the purchase complaints. Apple needs to review the whole presentation—not let users discover the difference after checkout. @Apple @AppStore @Google @googlechrome 🔗 Full investigation and evidence https://t.co/8IOGvVrbUT #MacAppStore #GoogleChrome #ConsumerProtection #ScamAwareness #AppStore

欺骗供应链 @patrickwardle 原文 ↗

只谈工具传播现象,无技术细节。

RT @ProbiusOfficial: ARTEX 本质上属于是因为某些原因出圈打破信息差了,再加上它本身就是一个极其优秀的工具 在安全领域优秀的工具很多 并不只有ARTEX一个 只是在AI爆发和AI自媒体渗透大众之前,网络安全领域相对于公众视野来讲一直是一个很小众和幕后的圈子,哪怕同为计算机的开发人员经常上GitHub 它也不会去关心GitHub上面的安全工具 直到....这个壁垒被自媒体和AI彻底打破😇

议题 @0xocdsec 原文 ↗

仅提升提交流程效率,与漏洞本身无关。

RT @TinySecEx: If you’re like me and have grown tired of the clunky portal system used by MSRC, try this browser extension —it’ll make your life easier.. https://t.co/p0BozyX93Z , Allowing cases to be opened in a new tab without requiring a re-login(I simply cannot fathom the design choices behind MSRC )

工具 @0xocdsec 原文 ↗

反映漏洞披露流程效率问题,无技术增量。

RT @soaj1664ashar: . @msftsecresponse " On average MSRC took 54 days — nearly two months — to tell us a report was a duplicate." + @secbughunter I wonder why it took you so long just to decide that this is a duplicate. This needs improvement. To me, “duplicate” means that you already had a report from someone else, which should have made the decision quick. Honestly, this feels seriously bad.

议题 @0xocdsec 原文 ↗

自建虚拟化平台被攻陷后攻击者可以长期隐匿驻留,Settra 勒索团伙已在使用。

Recorded Future's Lawrence Stowe & Jerrilee Plude look into GhostFrame, a post-exploitation framework that gives threat actors control over compromised VMware ESXi environments. This is used by TAG-218, a threat activity group linked to Settra ransomware. https://t.co/ctyDwTrDca https://t.co/ZXxCAhrQvN

云后渗透工具报告 @virusbtn 原文 ↗

自托管 Exchange 一旦失陷,邮件收集模块在磁盘上几乎不留痕。

ESET's Matías Porolli describes Turla’s Cereal toolset - featuring a novel Exchange email collection module, a unique C&C mechanism, & sophisticated .NET execution tricks - which was found in compromised governmental organizations in Kazakhstan & Romania. https://t.co/hnR9BMtkiO https://t.co/lLjLe7sArz

aptexchange后渗透报告 @virusbtn 原文 ↗

攻击形态从诱导执行升级为长期驻留,终端侧的检测规则需要更新。

RT @HuntressLabs: ClickFix has expanded beyond simple malware delivery into complex attack chains that establish persistence, support lateral movement, and enable reconnaissance and credential theft. 🖇️ Read our round-up of the latest ClickFix variations we’ve seen: https://t.co/Dpwf4hnd3l https://t.co/fiIC15m2In

钓鱼后渗透报告 @virusbtn 原文 ↗

混用系统原生工具后行为特征稀疏,可直接作为检测基线参考。

RT @FortiGuardLabs: Three infection chains with one objective: persistent access. ✅ Our team is tracing a BITTER campaign targeting South Asian organizations that blends native Windows tools, malicious shortcuts, and MSI installers to deploy #PapRAT, a custom RAT now evolving from C++ to Rust with expanded capabilities. 🔗 Follow for guidance and mitigation tips: https://t.co/vriz6887zJ #FortiGuardLabs

aptrat钓鱼 @virusbtn 原文 ↗

身份优先的云入侵路径,密码重置与 MFA 流程是首要阻断环节。

RT @MsftSecIntel: Microsoft tracks Storm-3350, a Brazil-focused activity cluster observed since October, primarily affecting organizations in financial services, payments, banking, and digital assets. Microsoft-observed activity has also affected healthcare, transportation, and logistics. Storm-3350 uses an identity-first cloud intrusion pattern that can begin with voice phishing (vishing) or identity-focused impersonation. In confirmed activity, the actor reset an account password through Microsoft’s legitimate self-service password-reset flow, signed in within minutes, completed the account’s existing SMS or voice multifactor authentication (MFA) challenge, and registered an actor-controlled Microsoft Authenticator method. This suggests an ability to intercept or receive codes sent to the registered phone number, potentially through social engineering involving employees’ personal phones. The new method provides persistent access without continued dependence on the victim or their phone. After account takeover, Storm-3350 conducts reconnaissance and high-volume cloud-data collection across Microsoft Graph, Microsoft 365, and Azure, identifying high-value identities, applications, sites, drives, mailboxes, and sensitive content. This activity combines browser access with Python- and Node-style automation, OAuth-authenticated curl activity, and Azure CLI enumeration. Storm-3350’s observed activity also included searching SharePoint for “AKIA” and accessing .key, .pem, and .ppk files. In Azure, the actor requested storage-account keys and application publishing profiles, which could enable access to additional cloud resources. Together, these actions point to possible routes beyond the compromised account. Defenders should investigate the full sequence rather than isolated events: password reset; immediate sign-in using existing MFA; authentication-method registration; anomalous token acquisition; Microsoft Graph and Azure control-plane reconnaissance; SharePoint, OneDrive, and mailbox discovery; sensitive-content searches; credential requests; and high-volume retrieval or download. Relevant detections include: - Suspicious Azure List Keys activities related to the List Keys operation - Anomalous token acquisition followed by high-volume SharePoint file activity - Graph API activity from a Microsoft threat intelligence IP address - Mass SharePoint or OneDrive file access after an MFA method change - Automated mass SharePoint/OneDrive file access via python-httpx - Suspicious mass download from risky or unusual session Microsoft tracks Storm-3350 separately within a broader financially motivated identity-compromise and cloud-data-theft ecosystem. Because related clusters share tactics, infrastructure patterns, and operational dependencies, no single IP, domain, or lure theme should be treated as conclusive actor attribution.

云身份钓鱼 @virusbtn 原文 ↗

住宅代理基础设施被拆除,相关代理服务的可用性与可信度将出现变化。

RT @nicter_jp: Googleが先日、レジデンシャルプロキシなどに悪用されていたXCPEボットネットに対する国際的なテイクダウン作戦を発表しました。日本からは警察庁も参加しています。 NICTでも以前からXCPEの調査・分析を進めており、その解析結果や国内における感染機器の観測状況をブログにまとめました。 https://t.co/8tACtaIOKx

僵尸网络代理事件 @virusbtn 原文 ↗

泄露材料中包含活跃 webshell,可作为失陷指标供防守方比对。

[DATA LEAK] "NEXUSEC" claims to sell Cambodia's NCSTI database on PwnForums. • 🇰🇭 Cambodia — NCSTI, a government science and technology council under MISTI. • 40 accounts allegedly exposed, including 7 high-privilege admins (1 superadmin) and 33 national talent profiles. • Credentials and RBAC role records across 24 administrative modules. • 30 policy and technology roadmap documents, plus a MinIO store of 455 objects (~490 MB). • Includes 6 PHP webshells and a claimed active command-execution backdoor. ✔ Posted on PwnForums by "NEXUSEC", with a downloadable archive linked for the dataset. The listing describes admin accounts tied to official misti gov kh addresses and a live webshell enabling unauthenticated remote code execution on the host. 📌 S2W note: The claimed active webshell and RCE backdoor point to ongoing access rather than a one-off dump, raising the risk of continued compromise. Follow @S2W_DailyThreat #DataLeak #Cambodia #NCSTI #S2W

数据泄露webshell事件 @S2W_DailyThreat 原文 ↗

可复用的 Windows 内部机制追踪方法,直接用于样本分析。

To learn Windows internals reversing, I trace Win32 functions in Ghidra end to end, including their kernel journey. TerminateProcess: - kernel32.dll - kernelbase.dll.TerminateProcess - ntdll.dll.NtTerminateProcess - ntoskrnl.exe.ZwTerminateProcess The rest of chain is on you ;) https://t.co/2zZnr0g27f

逆向windows议题 @papadelospapas_ 原文 ↗

RT @jpcert_ac: 「インシデント調査報告書チェックリスト」を公開しました。 本コンテンツは、セキュリティインシデント発生時に被害組織と調査会社の双方がスムーズに調査を進められることをサポートするものになっていますので、インシデント調査を外部委託する際はご利用ください。 https://t.co/a6o1AQRht8

应急响应报告 @virusbtn 原文 ↗

Hyper-V、故障转移集群与 DFS 共用 SPN,是域内取盘的新路径。

RT @unsigned_sh0rt: Quick blog on how iSCSI initiators/targets work in AD domains. If you can guess the right IQN you've got access to the disk. Haven't done the attack in prod but see the SPN shared with things like Hyper-V, Failover Clusters, and DFS.

域渗透横向移动存储 @0x64616e 原文 ↗

跨平台红队载荷构建的实操整理,基于 OSEP 课程内容。

RT @kalilinux: Building Windows C# Tools and Payloads from Kali Linux: One of our team members recently took (and passed!) OffSec’s advanced penetration testing course & exam, PEN-300/OSEP. Whilst it focuses a lot on Microsoft Windows, there were a… www.​kali.​org/blog/windows-csharp-payloads/ https://t.co/sGFwyMqFG2

工具红队载荷 @0x64616e 原文 ↗

附 AnyRun API 复现步骤,可直接转成 EDR 检测规则。

https://t.co/nNz09B6NGM next one from my blog! enjoy! a similar #technique is used by Turla Kazuar #malware and PHANTOMPULSE Thanks to @anyrun_app for API! ♥️ https://t.co/PpMhcbEqq5 #hacking #maldev #redteam #research #windows #injection #debugging #threatintel #cybersecurity https://t.co/XN25fclZDi

注入恶意软件检测 @cocomelonckz 原文 ↗

[INITIAL ACCESS] "dalxyon" claims to sell cloud dashboard access to a Saudi contractor on DarkForums. • 🇸🇦 Saudi Arabia • Adwaa Al Khalil Contracting Co., a construction and infrastructure firm. • Allegedly offers an active, verified cloud admin dashboard session. • ~240.3 GB claimed, including tender proposals and engineering documents. • Fixed price $500, payable in BTC, XMR, or USDT. ✔ Posted on DarkForums by "dalxyon"; the listing describes recent November 2026 archives covering closed technical and financial tenders plus project documentation for government facilities in Arabic, and points buyers to a Telegram contact. 📌 S2W note: The combination of an active admin session and government-facility tender data suggests both espionage and resale value to a buyer. Follow @S2W_DailyThreat #InitialAccess #DarkForums #SaudiArabia #AdwaaAlKhalil #S2W

云初始访问事件 @S2W_DailyThreat 原文 ↗

[INITIAL ACCESS] "marcustran87" claims to sell root access to a Vietnam corporate gateway on Breachforums_V9. • 🇻🇳 Vietnam — an enterprise VPN/router on a VNPT business line in Thai Nguyen. • Root shell plus 5 working VPN accounts allegedly offered. • Access purportedly reaches ~35 internal hosts and a small DMZ (NAS and a Windows server). • Static IP on a 100Mbps business line. • Listed at $180, open to offers, escrow only. ✔ Posted by "marcustran87" on Breachforums_V9, advertising long-running, live access to the box and the segment behind it. The actor requires escrow before any handover and offers a live demo through the middleman, accepting XMR, BTC, and USDT. 📌 S2W note: The escrow-only terms and offer of a live demo are presented to bolster the credibility of the claimed access. Follow @S2W_DailyThreat #InitialAccess #Vietnam #UndergroundForum #S2W

初始访问事件 @S2W_DailyThreat 原文 ↗

[DATA LEAK] "Sensitive2025" claims to sell AdiMudra banking data on PwnForums. • 🇮🇳 India • AdiMudra, an online financial services provider. • ~180,000 lines of alleged bank information offered for sale. ✔ Posted on PwnForums by "Sensitive2025", the listing advertises banking records and points buyers to a Telegram contact to start a deal. The actor references sample data and an escrow app to facilitate transactions. 📌 S2W note: The pairing of banking records with a dedicated contact channel and escrow offer points to clear resale intent. Follow @S2W_DailyThreat #DataLeak #Finance #India #UndergroundForum #S2W

数据泄露事件 @S2W_DailyThreat 原文 ↗

[DATA LEAK] "bibimarket" claims to leak Israeli MoD contract docs on Breachforums_V9. • 🇮🇱 Israel • Nemsysco, a maker of voice-analysis lie-detection and voice-identification systems. • Documents allegedly include Ministry of Defense export authorizations. • The actor references the LVA 6.5 lie detector among the exposed material. ✔ Posted on Breachforums_V9 by "bibimarket", who frames the leak as part of a larger campaign targeting Israeli officials and promotes a Telegram channel to distribute further material. 📌 S2W note: Per the post, the authorizations cover technology reportedly deployed at airports in Russia and Israel, giving the leak a defense-export sensitivity beyond a standard corporate document dump. Follow @S2W_DailyThreat #DataLeak #Nemsysco #Israel #Defense #S2W

数据泄露事件 @S2W_DailyThreat 原文 ↗

[DATA LEAK] "UNC6783" claims to leak Qatar National Bank data on DarkForums. • 🇶🇦 Qatar — Qatar National Bank (Q.P.S.C.), a commercial bank. • ~25M customer records allegedly exposed. • ~28K employee records also claimed. • Customer samples include names, PO box addresses, phone numbers, and nationality codes. • The actor provided a Telegram contact channel. ✔ Posted on DarkForums by "UNC6783", who attached sample customer rows to substantiate the claimed volume. The listing bundles personal and contact details across customers from multiple nationalities, with a Telegram channel for follow-up. 📌 S2W note: The combination of bulk customer PII and an open contact channel points to resale or extortion intent. Follow @S2W_DailyThreat #DataLeak #QatarNationalBank #Qatar #DarkWeb #S2W

数据泄露事件 @S2W_DailyThreat 原文 ↗

为可定制外观而拒绝升级系统,会让日常使用的手机长期暴露在已经修复的漏洞下。

RT @pwnlambda: Reminder: keeping phones on 27.0.1 to customize how they look is very dangerous as they’ll also be vulnerable to this. If you want a device to perform security research on and don’t use it for anything else it should be fine but daily driving one is begging for trouble.

移动端ios漏洞 @0xocdsec 原文 ↗

前沿模型的攻防能力评估被厂商准入条件卡住,第三方 AI 安全评测因此难以开展。

RT @mdp_sec: gemini 4 carbon dropped this morning and i can't bench it google gates the cyber side behind a program called fairwind our application has been pending for months every other lab let me test day one google is the only one where legit nobody i know has access @GoogleDeepMind who is actually in fairwind

llmai_agent事件 @0xocdsec 原文 ↗