CVE & CISA-KEV Catalog

vLLM is an inference and serving engine for large language models (LLMs). Prior to 0.23.1rc0, ll temperature validation gates use comparison operators (<, >), which silently evaluate to False for NaN and for positive Infinity in Python's IEEE 754 float semantics. Both values pass every guard and propagate to GPU sampling kernels, where they produce undefined behavior or CUDA errors that can crash the inference worker. This vulnerability is fixed in 0.23.1rc0.

Published 2026-06-22

A denial-of-service vulnerability exists in the WebSocket API due to insufficient validation and handling of JSON-based requests. A low-privileged authenticated attacker can send a specially crafted request that causes service disruption and may result in an unexpected device reboot.

Published 2026-06-16

Netty is a network application framework for development of protocol servers and clients. In netty-handler prior to versions 4.1.135.Final and 4.2.15.Final, an attacker can bypass IPv6 subnet rules due to an incorrect masking operation in IpSubnetFilterRule.compareTo(). Valid public IP addresses can bypass the restrictions. Versions 4.1.135.Final and 4.2.15.Final patch the issue.

Published 2026-06-11

The $_internalApplyOplogUpdate aggregation pipeline stage can be used to execute a document diff containing a malformed binary diff to return memory out-of-bounds or crash the server. $_internalApplyOplogUpdate can be executed by any authenticated user with access to the aggregate command.

Published 2026-06-09

When OIDC authentication is enabled in configuration, clients may set specific values in the "mechanism" parameter of the "authenticate" command that lead to server crash. The authenticate command is accessible to unauthenticated clients, leading to pre-auth denial-of-service in affected product configurations.

Published 2026-06-09

A flaw has been found in Boost Serialization up to 1.91. The impacted element is an unknown function. This manipulation causes improper validation of specified type of input. It is possible to initiate the attack remotely. The exploit has been published and may be used. The maintainer was notified on Aug 2025 and a disclosure deadline was set for 90 days. The maintainer acknowledged but postponed indefinitely citing time concerns. No patch is currently available and the disclosure deadline has expired.

Published 2026-06-07

In Arista’s EOS when in 802.1X mode, multi-auth unauthenticated hosts might be allowed access to a switch port if there exists an EAPOL capable device in the fallback VLAN.

Published 2026-06-04

Net::CIDR::Set versions through 0.20 for Perl did not validate IP addresses. The add method called the _encode method to parse addresses. If the addresses did not look like netmasks or network ranges, then they were assumed to single IP addresses and passed back to itself as a 32-bit or 128-bit netmask. If the argument was not a well-formed IP address, then this would lead to indefinite recursion. An attacker could use this to cause a denial of service.

Published 2026-06-04

Hono is a Web application framework that provides support for any JavaScript runtime. Prior to 4.12.21, the serialize() function in hono/cookie validates domain and path options against characters that corrupt Set-Cookie header syntax (;, \r, \n), but does not apply the same validation to sameSite and priority. An application that passes user-controlled input into either option may produce a Set-Cookie response header containing attacker-chosen additional attributes. This vulnerability is fixed in 4.12.21.

Published 2026-05-28

A local attacker can perform a confusion attack on the cfgparser via a specially crafted file on an USB stick leading to code execution. This can result in a total loss of confidentiality, integrity and availability.

Published 2026-05-27

A security vulnerability has been detected in fraillt bitsery up to 5.2.4. Affected is the function loadFromSharedState in the library include/bitsery/ext/std_smart_ptr.h. Such manipulation leads to improper validation of specified type of input. It is possible to launch the attack remotely. The exploit has been disclosed publicly and may be used. Upgrading to version 5.2.5 is able to address this issue. The name of the patch is 66d16516e24893bebc1c8af52bf2fe9ad0735061. Upgrading the affected component is advised.

Published 2026-05-26

Multiple flaws have been identified in `named` related to the handling of DNS messages whose CLASS is not Internet (`IN`) — for example, `CHAOS` or `HESIOD`, or DNS messages that specify meta-classes (`ANY` or `NONE`) in the question section. Specially crafted requests reaching the affected code paths — recursion, dynamic updates (`UPDATE`), zone change notifications (`NOTIFY`), or processing of `IN`-specific record types in non-`IN` data — can cause assertion failures in `named`. This issue affects BIND 9 versions 9.11.0 through 9.16.50, 9.18.0 through 9.18.48, 9.20.0 through 9.20.22, 9.21.0 through 9.21.21, 9.11.3-S1 through 9.16.50-S1, 9.18.11-S1 through 9.18.48-S1, and 9.20.9-S1 through 9.20.22-S1.

Published 2026-05-23

For Concrete CMS 9.5.0 and below, OAuth 2.0 Authorization-Code Handler Bypasses Account Status. A user with uIsActive=0 (suspended, banned, terminated employee) can still authenticate via OAuth and receive valid API tokens. The Concrete CMS security team gave this vulnerability a CVSS v.4.0 score of 2.3 with vector CVSS:4.0/AV:N/AC:L/AT:P/PR:L/UI:N/VC:L/VI:L/VA:N/SC:L/SI:L/SA:N. Thanks 0x4c616e for reporting.

Published 2026-05-21

An ACAP configuration file lacked sufficient input validation, which could allow command injection and potentially lead to privilege escalation. This vulnerability can only be exploited if the Axis device is configured to allow the installation of unsigned ACAP applications, and if an attacker convinces the victim to install a malicious ACAP application.

Published 2026-05-12

NEMU (OpenXiangShan/NEMU) before v2025.12.r2 contains an improper instruction-validation flaw in its RISC-V Vector (RVV) decoder. The decoder does not correctly validate the funct3 field when decoding vsetvli/vsetivli/vsetvl, allowing certain invalid OP-V instruction encodings to be misinterpreted and executed as vset* configuration instructions rather than raising an illegal-instruction exception. This can be exploited by providing crafted RISC-V binaries to cause incorrect trap behavior, architectural state corruption/divergence, and potential denial of service in systems that rely on NEMU for correct execution or sandboxing.

Published 2026-04-20

Impact: Fastify applications using schema.body.content for per-content-type body validation can have validation bypassed entirely by prepending a space to the Content-Type header. The body is still parsed correctly but schema validation is skipped. This is a regression introduced in fastify >= 5.3.2 by the fix for CVE-2025-32442 Patches: Upgrade to fastify v5.8.5 or later. Workarounds: None. Upgrade to the patched version.

Published 2026-04-15

Versions of the package jsrsasign before 11.1.1 are vulnerable to Infinite loop via the bnModInverse function in ext/jsbn2.js when the BigInteger.modInverse implementation receives zero or negative inputs, allowing an attacker to hang the process permanently by supplying such crafted values (e.g., modInverse(0, m) or modInverse(-1, m)).

Published 2026-03-23

SpotAuditor 5.2.6 contains a denial of service vulnerability in the registration dialog that allows local attackers to crash the application by supplying an excessively long string in the Name field. Attackers can paste a buffer of 300 repeated characters into the Name input during registration to trigger an application crash.

Published 2026-03-22

A flaw was found in Keycloak. Keycloak's Security Assertion Markup Language (SAML) broker endpoint does not properly validate encrypted assertions when the overall SAML response is not signed. An attacker with a valid signed SAML assertion can exploit this by crafting a malicious SAML response. This allows the attacker to inject an encrypted assertion for an arbitrary principal, leading to unauthorized access and potential information disclosure.

Published 2026-03-18

Mattermost versions 11.3.x <= 11.3.0, 11.2.x <= 11.2.2, 10.11.x <= 10.11.10 fail to handle incorrectly reported array lengths which allows malicious user to cause OOM errors and crash the server via sending corrupted msgpack frames within websocket messages to calls plugin. Mattermost Advisory ID: MMSA-2025-00537

Published 2026-03-16

Mattermost versions 11.3.x <= 11.3.0, 11.2.x <= 11.2.2, 10.11.x <= 10.11.10 fail to properly validate User-Agent header tokens which allows an authenticated attacker to cause a request panic via a specially crafted User-Agent header. Mattermost Advisory ID: MMSA-2026-00586

Published 2026-03-16

A vulnerability in the Intermediate System-to-Intermediate System (IS-IS) multi-instance routing feature of Cisco IOS XR Software could allow an unauthenticated, adjacent attacker to cause the IS-IS process to restart unexpectedly. This vulnerability is due to insufficient input validation of ingress IS-IS packets. An attacker could exploit this vulnerability by sending crafted IS-IS packets to an affected device after forming an adjacency. A successful exploit could allow the attacker to cause the IS-IS process to restart unexpectedly, resulting in a temporary loss of connectivity to advertised networks and a denial of service (DoS) condition. Note: The IS-IS protocol is a routing protocol. To exploit this vulnerability, an attacker must be Layer 2-adjacent to the affected device and

Published 2026-03-11

Improper validation of specified type of input in SQL Server allows an authorized attacker to elevate privileges over a network.

Published 2026-03-10

Improper validation of specified type of input in Windows Ancillary Function Driver for WinSock allows an authorized attacker to elevate privileges locally.

Published 2026-03-10

TSPortal is the WikiTide Foundation’s in-house platform used by the Trust and Safety team to manage reports, investigations, appeals, and transparency work. Prior to version 30, conversion of empty strings to null allows disguising DPA reports as genuine self-deletion reports. This issue has been patched in version 30.

Published 2026-03-06

Missing validation of type of input in PostgreSQL intarray extension selectivity estimator function allows an object creator to execute arbitrary code as the operating system user running the database. Versions before PostgreSQL 18.2, 17.8, 16.12, 15.16, and 14.21 are affected.

Published 2026-02-12

Improper validation of type "oidvector" in PostgreSQL allows a database user to disclose a few bytes of server memory. We have not ruled out viability of attacks that arrange for presence of confidential information in disclosed bytes, but they seem unlikely. Versions before PostgreSQL 18.2, 17.8, 16.12, 15.16, and 14.21 are affected.

Published 2026-02-12

Axios is a promise based HTTP client for the browser and Node.js. Prior to versions 0.30.3 and 1.13.5, the mergeConfig function in axios crashes with a TypeError when processing configuration objects containing __proto__ as an own property. An attacker can trigger this by providing a malicious configuration object created via JSON.parse(), causing complete denial of service. This vulnerability is fixed in versions 0.30.3 and 1.13.5.

Published 2026-02-09

A vulnerability in the text rendering subsystem of Cisco TelePresence Collaboration Endpoint (CE) Software and Cisco RoomOS Software could allow an unauthenticated, remote attacker to cause a denial of service (DoS) condition on an affected device. This vulnerability is due to insufficient validation of input received by an affected device. An attacker could exploit this vulnerability by getting the affected device to render crafted text, for example, a crafted meeting invitation. As indicated in the CVSS score, no user interaction is required, such as accepting the meeting invitation. A successful exploit could allow the attacker to cause the affected device to reload, resulting in a DoS condition.

Published 2026-02-04

Improper validation of specified type of input in M365 Copilot allows an unauthorized attacker to disclose information over a network.

Published 2026-01-22

Vulnerability in the Oracle Java SE, Oracle GraalVM for JDK, Oracle GraalVM Enterprise Edition product of Oracle Java SE (component: AWT, JavaFX). Supported versions that are affected are Oracle Java SE: 8u471, 8u471-b50, 8u471-perf, 11.0.29, 17.0.17, 21.0.9, 25.0.1; Oracle GraalVM for JDK: 17.0.17 and 21.0.9; Oracle GraalVM Enterprise Edition: 21.3.16. Easily exploitable vulnerability allows unauthenticated attacker with network access via multiple protocols to compromise Oracle Java SE, Oracle GraalVM for JDK, Oracle GraalVM Enterprise Edition. Successful attacks require human interaction from a person other than the attacker and while the vulnerability is in Oracle Java SE, Oracle GraalVM for JDK, Oracle GraalVM Enterprise Edition, attacks may significantly impact additional products

Published 2026-01-20

Meshtastic is an open source mesh networking solution. The Meshtastic firmware (starting from version 2.5) introduces asymmetric encryption (PKI) for direct messages, but when the `pki_encrypted` flag is missing, the firmware silently falls back to legacy AES-256-CTR channel encryption. This was an intentional decision to maintain backwards compatibility. However, the end-user applications, like Web app, iOS/Android app, and applications built on top of Meshtastic using the SDK, did not have a way to differentiate between end-to-end encrypted DMs and the legacy DMs. This creates a downgrade attack path where adversaries who know a shared channel key can craft and inject spoofed direct messages that are displayed as if they were PKC encrypted. Users are not given any feedback of whether a d

Published 2025-12-29

Mattermost versions 11.0.x <= 11.0.4, 10.12.x <= 10.12.2, 10.11.x <= 10.11.6 fail to check WebSocket request field for proper UTF-8 format, which allows attacker to crash Calls plug-in via sending malformed request.

Published 2025-12-17

Mattermost versions 10.11.x <= 10.11.6 and Mattermost GitHub plugin versions <=2.4.0 fail to validate plugin bot identity in reaction forwarding which allows attackers to hijack the GitHub reaction feature to make users add reactions to arbitrary GitHub objects via crafted notification posts.

Published 2025-12-17

An unauthorised attacker within bluetooth range may use an improper validation during the BLE connection request to deadlock the affected devices.

Published 2025-12-10

Fluent Bit in_http, in_splunk, and in_elasticsearch input plugins fail to sanitize tag_key inputs. An attacker with network access or the ability to write records into Splunk or Elasticsearch can supply tag_key values containing special characters such as newlines or ../ that are treated as valid tags. Because tags influence routing and some outputs derive filenames or contents from tags, this can allow newline injection, path traversal, forged record injection, or log misrouting, impacting data integrity and log routing.

Published 2025-12-07

In KDE Connect before 1.33.0 on Android, malicious device IDs (sent via broadcast UDP) could cause an application crash.

Published 2025-12-05

In Modem, there is a possible system crash due to a logic error. This could lead to remote denial of service, if a UE has connected to a rogue base station controlled by the attacker, with no additional execution privileges needed. User interaction is not needed for exploitation. Patch ID: MOLY01673749; Issue ID: MSV-4643.

Published 2025-12-02

An issue was discovered in Free5GC v4.0.0 and v4.0.1 allowing an attacker to cause a denial of service via the Nudm_SubscriberDataManagement API.

Published 2025-11-24

An unauthenticated remote attacker can send a specially crafted Modbus read command to the device which leads to a denial of service.

Published 2025-11-24

The VAPIX API port.cgi did not have sufficient input validation, which may result in process crashes and impact usability. This vulnerability can only be exploited after authenticating with a viewer- operator- or administrator-privileged service account.

Published 2025-11-11

An ACAP configuration file has improper permissions and lacks input validation, which could potentially lead to privilege escalation. This vulnerability can only be exploited if the Axis device is configured to allow the installation of unsigned ACAP applications, and if an attacker convinces the victim to install a malicious ACAP application.

Published 2025-11-11

ACAP applications can gain elevated privileges due to improper input validation, potentially leading to privilege escalation. This vulnerability can only be exploited if the Axis device is configured to allow the installation of unsigned ACAP applications, and if an attacker convinces the victim to install a malicious ACAP application.

Published 2025-11-11

An ACAP configuration file lacked sufficient input validation, which could allow for arbitrary code execution. This vulnerability can only be exploited if the Axis device is configured to allow the installation of unsigned ACAP applications, and if an attacker convinces the victim to install a malicious ACAP application.

Published 2025-11-11

Improper validation of specified type of input in Windows Authentication Methods allows an authorized attacker to elevate privileges locally.

Published 2025-10-14

Improper validation of specified type of input in Windows Authentication Methods allows an authorized attacker to elevate privileges locally.

Published 2025-10-14

Improper validation of specified type of input in Windows Authentication Methods allows an authorized attacker to elevate privileges locally.

Published 2025-10-14

Improper validation of specified type of input in Windows Local Session Manager (LSM) allows an authorized attacker to deny service over a network.

Published 2025-10-14

Improper validation of specified type of input in Windows Local Session Manager (LSM) allows an authorized attacker to deny service over a network.

Published 2025-10-14

Improper validation of specified type of input in Windows Local Session Manager (LSM) allows an authorized attacker to deny service over a network.

Published 2025-10-14