2026-09-17 归档

今日共收录 2425 条安全情报，包含 1381 个 CVE，651 篇安全通告，以及 100 篇研究论文。

安全公告2026-09-16 发布INFO

Builder.io Gen2 SDKs through versions 5.2.11 and 0.25.13 contain a prototype pollution...

CVE-2026-92781

Builder.io Gen2 SDKs through versions 5.2.11 and 0.25.13 contain a prototype pollution vulnerability in the unflatten he

历史摘要 · 未重新核验 · GitHub Advisory

#advisory #cwe-1321 #severity:medium

安全公告2026-09-16 发布INFO

KnowStreaming through 3.4.1 fails to enforce role-based access control on REST API endpoints,...

CVE-2026-92780

KnowStreaming through 3.4.1 fails to enforce role-based access control on REST API endpoints, allowing any authenticated

历史摘要 · 未重新核验 · GitHub Advisory

#advisory #cwe-862 #severity:high

安全公告2026-09-16 发布INFO

Wiki.js through 2.5.314 contains a server-side request forgery vulnerability in the Image...

CVE-2026-92775

Wiki.js through 2.5.314 contains a server-side request forgery vulnerability in the Image Prefetch renderer that fetches

历史摘要 · 未重新核验 · GitHub Advisory

#advisory #cwe-918 #severity:high

安全公告2026-09-16 发布INFO

Rundeck through 6.2.1 fails to properly authorize the importConfig and importNodesSources...

CVE-2026-92763

Rundeck through 6.2.1 fails to properly authorize the importConfig and importNodesSources parameters in the project arch

历史摘要 · 未重新核验 · GitHub Advisory

#advisory #cwe-862 #severity:high

安全公告2026-09-16 发布INFO

WebVirtCloud fails to properly validate permission flags in UserInstance grants, allowing view...

CVE-2026-92761

WebVirtCloud fails to properly validate permission flags in UserInstance grants, allowing view-only users to perform pri

历史摘要 · 未重新核验 · GitHub Advisory

#advisory #cwe-862 #severity:high

安全公告2026-09-16 发布INFO

A vulnerability in Cisco Identity Services Engine (ISE) and Cisco ISE Passive Identity Connector ...

CVE-2026-76450

A vulnerability in Cisco Identity Services Engine (ISE) and Cisco ISE Passive Identity Connector (ISE-PIC) could allow a

历史摘要 · 未重新核验 · GitHub Advisory

#advisory #cwe-564 #severity:medium

安全公告2026-09-16 发布INFO

djust has Cross-Site Request Forgery on the Server-Sent-Events transport: a cross-origin page can drive a victim-authenticated SSE session

CVE-2026-61593

### Impact The SSE client→server POST endpoints are `@csrf_exempt` and the SSE GET stream endpoint had no Origin check,

历史摘要 · 未重新核验 · GitHub Advisory

#advisory #pip #cwe-352 #severity:high

安全公告2026-09-16 发布INFO

LMdeploy has Remote Code Execution by Pickle Deserialization via zmq_rpc.call_and_response() in InterLM/lmdeploy

CVE-2025-59953

### Description The LMdeploy implements an rpc server (AsyncRPCServer in zmq_rpc.py) for supporting the RPC communicati

历史摘要 · 未重新核验 · GitHub Advisory

#advisory #pip #cwe-502 #severity:critical

安全公告2026-09-16 发布INFO

@zereight/mcp-gitlab: Unauthenticated arbitrary file read via `upload_markdown` enables PAT exfiltration and full account takeover

CVE-2026-61560

### Summary The SSE transport mode (`SSE=true`) exposes all MCP tools without any authentication. The `upload_markdown`

历史摘要 · 未重新核验 · GitHub Advisory

#advisory #npm #cwe-22 #severity:critical

安全公告2026-09-15 发布INFO

@zereight/mcp-gitlab Vulnerable to Server-Side Request Forgery

CVE-2026-61559

# Server-Side Request Forgery via X-GitLab-API-URL Header Allows Credential Theft ## Affected - **Repository:** `zerei

历史摘要 · 未重新核验 · GitHub Advisory

#advisory #npm #cwe-918 #severity:critical

安全公告2026-09-15 发布INFO

Http4s: Ember HTTP/2 buffers a frame's declared payload before checking SETTINGS_MAX_FRAME_SIZE

CVE-2026-88975

### Summary An unauthenticated peer can make Ember's HTTP/2 read loop hold 16 MiB of a single frame in memory on a conne

历史摘要 · 未重新核验 · GitHub Advisory

#advisory #maven #cwe-400 #severity:high

学术论文2026-09-14 发布INFO

RuleAutoPilot: Synthesizing Deployable Suricata Rules from Network Traffic

论文提出 RuleAutoPilot，一个端到端的智能体（agentic）框架，可直接从恶意软件网络流量（PCAP）中合成可部署的 Suricata 检测规则，且不依赖任何事先准备好的威胁情报。研究背景是：以 Suricata 为代表的基于

历史摘要 · 未重新核验 · arXiv cs.CR

#suricata #ids #llm-agent #detection-engineering #network-security #rule-generation

学术论文2026-09-16 发布INFO

Autonomy in Check: Governor-Mediated Adaptive Security at the Edge

该论文研究的是网络边缘场景下自适应安全（adaptive security）所面临的一类新型信任边界问题。随着边缘网络的安全决策越来越依赖自动化规划器（planner），包括基于规则的控制器、通过机器学习得到的策略，以及近期出现的 LLM

历史摘要 · 未重新核验 · arXiv cs.CR

#cs.AI #agent-security #llm #ebpf #edge-security #policy-enforcement

学术论文2026-09-16 发布INFO

PentestChain: A Cost-Aware, MCP-Orchestrated Framework for Automated Penetration Testing with Free-Tier LLMs

该论文针对「AI 驱动的自动化渗透测试成本过高」这一现实问题提出 PentestChain 框架。作者指出，虽然 GPT-4 等前沿商用模型已被证明可执行自动化渗透测试，但按次计费的 token 成本使持续、自动化测试对预算有限的中小组织而

历史摘要 · 未重新核验 · arXiv cs.CR

#cs.AI #cs.NI #llm #agent-security #mcp #penetration-testing

学术论文2026-09-15 发布MEDIUM

InceptionRAG: Stealthy Poisoning Attack Against Retrieval-Augmented Generation

本文研究检索增强生成（RAG）系统的语料投毒威胁。RAG 通过引入外部知识提升大语言模型（LLM）的事实性与时效性，但已有研究证明其知识库可被投毒。现有攻击大多属于“单点显式注入”范式：完整的恶意载荷被封装在单篇文档内，语义上可被识别，因此

历史摘要 · 未重新核验 · arXiv cs.CR

#rag-security #llm #data-poisoning #adversarial-ml #multi-hop-reasoning #retrieval-security

漏洞预警2026-09-16 发布CRITICAL

In the Linux kernel, the following vulnerability has been resolved: ceph: lock mutex in ceph_mds_check_access() MDS session OPEN handling replaces mdsc->s_cap_auths under mdsc...

CVE-2026-89990

该漏洞位于 Linux 内核的 Ceph 文件系统客户端代码中，具体函数为 ceph_mds_check_access()。Ceph 客户端在 MDS 会话被重新打开（session OPEN）时，会在持有 mdsc->mutex 的前提下

历史摘要 · 未重新核验 · NVD

#cve

威胁情报2026-09-17 发布INFO

Cloud & Application Security

历史摘要 · 未重新核验 · CrowdStrike

#apt #crowdstrike #edr

威胁情报2026-09-17 发布INFO

From Scanner to Stealer: Inside the trivy-action Supply Chain Compromise

历史摘要 · 未重新核验 · CrowdStrike

#apt #crowdstrike #edr

威胁情报2026-09-17 发布INFO

Next-Gen Identity Security

历史摘要 · 未重新核验 · CrowdStrike

#apt #crowdstrike #edr

威胁情报2026-09-17 发布INFO

CrowdStrike Expands Identity Leadership with OpenID and IDPro

历史摘要 · 未重新核验 · CrowdStrike

#apt #crowdstrike #edr