2026-09-23 归档

今日共收录 1513 条安全情报,包含 580 个 CVE,548 篇安全通告,以及 93 篇研究论文。

← 返回归档列表
安全公告2026-09-22 发布INFO

FastGPT safe axios SSRF guard still allows DNS rebinding TOCTOU on protected outbound requests

CVE-2026-84301

FastGPT is an open-source LLM platform for building AI applications on a knowledge base. Prior to 4.15.2, the safe Axios

历史摘要 · 未重新核验 · OSV.dev

安全公告2026-09-22 发布INFO

Cloudreve: Privilege Scope Bypass: State-Mutating Admin Operations Accessible via Read-Only OAuth Scope

CVE-2026-77637

Cloudreve is a self-hosted file management and sharing system. Prior to 4.18.0, tool.GET("wopi") and tool.POST("mail") i

历史摘要 · 未重新核验 · OSV.dev

安全公告2026-09-22 发布INFO

MISP Collections: Missing Authorization Check for Sharing Group on PUT Request in collections/add

CVE-2026-95671

In MISP, the CollectionsController add() method enforced the sharing-group usability authorization check and element cap

历史摘要 · 未重新核验 · OSV.dev

安全公告2026-09-23 发布INFO

Ivanti Endpoint Manager Mobile代码注入漏洞

该条目来自知道创宇 / Seebug 漏洞库,标题为「Ivanti Endpoint Manager Mobile 代码注入漏洞」。Ivanti Endpoint Manager Mobile(EPMM,前身为 MobileIron Cor

历史摘要 · 未重新核验 · 知道创宇 / Seebug

安全公告2026-09-23 发布INFO

Fortigate Firewalls - Remote Code Execution (EGREGIOUSBLUNDER)

Fortigate Firewalls - Remote Code Execution (EGREGIOUSBLUNDER)

历史摘要 · 未重新核验 · 知道创宇 / Seebug

安全公告2026-09-22 发布INFO

Hatchet DurableTask WorkerStatus gRPC resolves caller-supplied durable-task UUIDs via ListSatisfiedEntries with no tenant_id filter

CVE-2026-88978

# Cross-tenant disclosure risk on `DurableTask` bidi RPC This is a low-severity, low-risk cross-tenant data exposure vu

历史摘要 · 未重新核验 · GitHub Advisory

安全公告2026-09-22 发布INFO

Hatchet: Cross-tenant durable callback payload disclosure in Hatchet V1 Dispatcher

CVE-2026-84298

## Summary The V1 `DurableTask` stream handler registers worker-supplied durable task external IDs in an in-memory call

历史摘要 · 未重新核验 · GitHub Advisory

安全公告2026-09-22 发布INFO

Cloudreve: Privilege Scope Bypass: State-Mutating Admin Operations Accessible via Read-Only OAuth Scope

CVE-2026-77637

### Summary There is a privilege scope bypass in Cloudreve's admin API where two endpoints that mutate server state are

历史摘要 · 未重新核验 · GitHub Advisory

安全公告2026-09-22 发布INFO

lightrag-hku: Stored Cross-Site Scripting (XSS) in the LightRAG WebUI chat/answer renderer via ingested content

CVE-2026-86062

### Summary The LightRAG WebUI renders assistant/answer chat content as **raw HTML** — `react-markdown` is configured wi

历史摘要 · 未重新核验 · GitHub Advisory

安全公告2026-09-22 发布INFO

lightrag-hku: SSRF via IPv6-transition address bypass (NAT64, IPv4-compatible, 6to4) of the native-markdown image-download guard

CVE-2026-85740

## Summary LightRAG's native markdown parser downloads external images referenced by an uploaded markdown or textpack d

历史摘要 · 未重新核验 · GitHub Advisory

安全公告2026-09-22 发布INFO

lightrag-hku: No Rate Limiting on /login Endpoint Allows Brute-Force Attacks

CVE-2026-85734

### Summary The POST /login endpoint has no rate limiting, account lockout, or delay on failed attempts. An attacker can

历史摘要 · 未重新核验 · GitHub Advisory

安全公告2026-09-22 发布INFO

KubeEdge: Unbounded allocation in viaduct packer enables authenticated remote DoS against CloudHub

CVE-2026-62370

## Summary KubeEdge CloudHub uses the viaduct packer to decode messages received from connected peers. The packer reads

历史摘要 · 未重新核验 · GitHub Advisory

威胁情报2026-09-23 发布INFO

从一个辅助 API 看资源边界:CVE-2026-71486(vLLM)

CVE-2026-71486

先知社区发布一篇针对 vLLM(主流大语言模型推理与在线服务框架)的技术分析文章,标题为《从一个辅助 API 看资源边界:CVE-2026-71486(vLLM)》,讨论编号为 CVE-2026-71486 的安全问题。文章以 vLLM 对

历史摘要 · 未重新核验 · 先知社区

威胁情报2026-09-22 发布INFO

D-Link warns of max severity zero-day bug in DIR-822A routers

CVE-2026-86296

D-Link 发布安全公告,警告其已停止支持的旧款 DIR-822A 双频 Wi-Fi 路由器存在一个最高严重等级(Max Severity)的零日漏洞,编号 CVE-2026-86296。该漏洞目前没有官方补丁,且公开的漏洞验证代码(Po

历史摘要 · 未重新核验 · BleepingComputer

安全公告2026-09-23 发布INFO

IBM DataStage on Cloud Pak for Data 5.4.0.0 could allow a remote authenticated attacker to...

CVE-2026-16346

IBM DataStage on Cloud Pak for Data 5.4.0.0 could allow a remote authenticated attacker to execute arbitrary commands du

历史摘要 · 未重新核验 · GitHub Advisory

安全公告2026-09-23 发布INFO

Citrix NetScaler 内存泄漏(CVE-2025-5777)

CVE-2025-5777

Citrix NetScaler 内存泄漏(CVE-2025-5777)

历史摘要 · 未重新核验 · 知道创宇 / Seebug

威胁情报2026-09-23 发布INFO

CrowdStrike Announces Continuous Identity for AI Agents

本文为 CrowdStrike 于 2026-09-23 发布的官方产品公告,标题为《CrowdStrike 为 AI Agent 推出 Continuous Identity(持续身份)》。从标题与厂商定位来看,该能力面向的是企业环境中快

历史摘要 · 未重新核验 · CrowdStrike

学术论文2026-09-19 发布INFO

SelfOp: An Optimization Algorithm for Self-Improving Security Agents

LLM 智能体正被用于漏洞发现、漏洞复现与补丁生成等安全任务,但在模型层面提升其能力通常依赖专家演示或可计算奖励,而安全任务恰恰很少具备这两点:执行轨迹昂贵、失败原因难以诊断、奖励稀疏且难以自动判定。因此研究重心转向执行框架(harness

历史摘要 · 未重新核验 · arXiv cs.CR

安全公告2026-09-22 发布INFO

Stable Channel Update for Desktop

CVE-2026-95291 · CVE-2026-95301 · CVE-2026-95281 · CVE-2026-95274 · CVE-2026-95329 · CVE-2026-95373 · CVE-2026-95313 · CVE-2026-95283 · CVE-2026-95277 · CVE-2026-95338 · CVE-2026-95350 · CVE-2026-95318 · CVE-2026-95298 · CVE-2026-95322 · CVE-2026-95284 · CVE-2026-95315 · CVE-2026-95293 · CVE-2026-95355 · CVE-2026-95356 · CVE-2026-95339 · CVE-2026-95357 · CVE-2026-95282 · CVE-2026-95286 · CVE-2026-95348 · CVE-2026-95335 · CVE-2026-95372 · CVE-2026-95324 · CVE-2026-95310 · CVE-2026-95349

The Chrome team is delighted to announce the promotion of Chrome 154 to the stable channel for Windows, Mac and Linux. T

历史摘要 · 未重新核验 · Google Chrome Releases

安全公告2026-09-22 发布INFO

LightRAG: No Rate Limiting on /login Endpoint Allows Brute-Force Attacks

CVE-2026-85734

LightRAG provides simple and fast retrieval-augmented generation. Prior to 1.5.5, the POST /login endpoint in lightrag/a

历史摘要 · 未重新核验 · OSV.dev