Lucene search
+L

11722 matches found

Github Security Blog
Github Security Blog
•added 2026/06/03 3:30 p.m.•24 views

huggingface/transformers: Arbitrary Code Execution During Model Initialization in the LightGlue Model Loading Path

A vulnerability in the LightGlue model loading path of huggingface/transformers version 5.2.0 allows an attacker-controlled model repository to execute arbitrary code during model initialization. The issue arises because the trustremotecode parameter, intended to prevent remote code execution, is...

9.6CVSS7.4AI score0.00936EPSS
SaveExploits1References11Affected Software1
PyPA
PyPA
•added 2026/06/03 2:16 p.m.•28 views

PYSEC-2026-2290

A vulnerability in the LightGlue model loading path of huggingface/transformers version 5.2.0 allows an attacker-controlled model repository to execute arbitrary code during model initialization. The issue arises because the trustremotecode parameter, intended to prevent remote code execution, is...

9.6CVSS7.4AI score0.00936EPSS
SaveExploits1References8Affected Software1
NVD
NVD
•added 2026/06/03 2:16 p.m.•33 views

CVE-2026-5241

A vulnerability in the LightGlue model loading path of huggingface/transformers version 5.2.0 allows an attacker-controlled model repository to execute arbitrary code during model initialization. The issue arises because the trustremotecode parameter, intended to prevent remote code execution, is...

9.6CVSS0.00936EPSS
SaveExploits1References9
CVE
CVE
•added 2026/06/03 12:33 p.m.•148 views

CVE-2026-5241

Technical details (affected products, versions, fixes, or exploit specifics) are not publicly available in the provided connected documents. Monitor for updates from vendors and security advisories.

9.6CVSS8.4AI score0.00936EPSS
SaveExploits1References9Affected Software1
OSV
OSV
•added 2026/06/03 12:33 p.m.•18 views

CVE-2026-5241 Policy Bypass in LightGlue Nested Config Resolution in huggingface/transformers

A vulnerability in the LightGlue model loading path of huggingface/transformers version 5.2.0 allows an attacker-controlled model repository to execute arbitrary code during model initialization. The issue arises because the trustremotecode parameter, intended to prevent remote code execution, is...

7.7CVSS7.9AI score
SaveExploits0References11
EUVD
EUVD
•added 2026/06/03 12:33 p.m.•27 views

EUVD-2026-34084

A vulnerability in the LightGlue model loading path of huggingface/transformers version 5.2.0 allows an attacker-controlled model repository to execute arbitrary code during model initialization. The issue arises because the trustremotecode parameter, intended to prevent remote code execution, is...

8CVSS7.9AI score0.00936EPSS
SaveExploits1References2
attackerkb
attackerkb
•added 2026/06/03 12:33 p.m.•23 views

CVE-2026-5241

A vulnerability in the LightGlue model loading path of huggingface/transformers version 5.2.0 allows an attacker-controlled model repository to execute arbitrary code during model initialization. The issue arises because the trustremotecode parameter, intended to prevent remote code execution, is...

8CVSS7.9AI score0.00936EPSS
SaveExploits1References3
RedhatCVE
RedhatCVE
•added 2026/06/03 10:01 a.m.•28 views

CVE-2026-24088

Cryptographic Issue while processing a specific partition which allows unauthorized write access to load a customized bootloader...

8.2CVSS5.8AI score0.00071EPSS
SaveExploits2References1
EUVD
EUVD
•added 2026/06/03 9:39 a.m.•24 views

EUVD-2026-34069

ZDRES-232: resolveProxyClass Not Overridden - acceptMatchers Filter Bypass via java.lang.reflect.Proxy Assessment: Fully addressed. When the serialised stream contains a TCPROXYCLASSDESC the marker for a java.lang.reflect.Proxy , JDK’s ObjectInputStream.readProxyDesc is dispatched. JDK then calls...

9.8CVSS5.8AI score0.00504EPSS
SaveExploits0References1
Vulnrichment
Vulnrichment
•added 2026/06/03 9:39 a.m.•18 views

CVE-2026-47065 Apache MINA: Critical Deserialization Allow-list Bypass via resolveProxyClass - ZDRES-232

ZDRES-232: resolveProxyClass Not Overridden - acceptMatchers Filter Bypass via java.lang.reflect.Proxy Assessment: Fully addressed. When the serialised stream contains a TCPROXYCLASSDESC the marker for a java.lang.reflect.Proxy , JDK’s ObjectInputStream.readProxyDesc is dispatched. JDK then calls...

9.8CVSS5.8AI score0.00504EPSS
SaveExploits0References1
GithubExploit
GithubExploit
•added 2026/06/03 8:50 a.m.•47 views

Exploit for Improper Privilege Management in Microsoft

CVE-2026-50343: InstallService StaticPluginMap EoP Standard U...

7.8CVSS7.4AI score0.00298EPSS
SaveExploits1
Positive Technologies
Positive Technologies
•added 2026/06/03 12:00 a.m.•38 views

PT-2026-45946

Name of the Vulnerable Software and Affected Versions huggingface/transformers version 5.2.0 Description A flaw in the LightGlue model loading path allows an attacker-controlled model repository to execute arbitrary code during model initialization. The issue occurs because the trust remote code...

9.6CVSS8AI score0.00936EPSS
SaveExploits1References23
Positive Technologies
Positive Technologies
•added 2026/06/03 12:00 a.m.•40 views

PT-2026-46099

Summary Using CookieJar.load with untrusted input may allow arbitrary code execution. Impact Most applications using this function will be doing so with the user's own data, so this is unlikely to affect many applications. Workaround If an application does allow attacker controlled files to be...

6.4CVSS6.1AI score0.00495EPSS
SaveExploits0References5
Positive Technologies
Positive Technologies
•added 2026/06/03 12:00 a.m.•29 views

PT-2026-45986

Name of the Vulnerable Software and Affected Versions Cisco Finesse affected versions not specified Description Insufficient validation of user-supplied input for HTTP requests allows an unauthenticated remote attacker to load arbitrary files from remote locations into an active user session. An...

6.1CVSS6.1AI score0.0018EPSS
SaveExploits0References4
CNNVD
CNNVD
•added 2026/06/03 12:00 a.m.•30 views

Cisco Finesse 安全漏洞

Cisco Finesse is a call center management software developed by the American company Cisco. There is a security vulnerability in Cisco Finesse, which stems from insufficient validation of HTTP request inputs provided to users. This vulnerability could allow unauthorized remote attackers to load...

6.1CVSS5.4AI score0.0018EPSS
SaveExploits0References1
BDU FSTEC
BDU FSTEC
•added 2026/06/03 12:00 a.m.•11 views

The vulnerability of the Azure Orbital Spatio deployment and management service, a software platform of Microsoft Azure, allows a attacker to execute arbitrary code.

The vulnerability of the Azure Orbital Spatio deployment and management service of the Microsoft Azure software platform is related to the unlimited loading of dangerous files. Exploiting this vulnerability could allow a malicious actor to execute arbitrary code remotely...

10CVSS6AI score0.00974EPSS
SaveExploits0References2
attackerkb
attackerkb
•added 2026/06/02 7:17 p.m.•12 views

CVE-2021-4478

Dräger CC-Vision Basic before 7.5.3 and Dräger CC-Vision E-Cal before 7.2.5.0 contain an out-of-bounds write vulnerability when loading .gdt files. A crafted .gdt file can trigger a buffer overflow during file parsing, allowing an attacker to crash the application or execute malicious code on the...

8.3CVSS6.3AI score0.00122EPSS
SaveExploits0References2
OSV
OSV
•added 2026/06/02 6:31 p.m.•19 views

GHSA-M3V4-V5GX-7WF5 OpenMed vulnerable to remote code injection through privacy-filter model loading path

OpenMed before 1.5.2 contains a remote code execution vulnerability in the PII privacy-filter model loading path. The privacy-filter dispatcher used broad substring matching on the user-supplied modelname parameter, allowing a value such as attacker/foo-privacy-filter-bar to route through a path...

9.8CVSS6.4AI score0.01288EPSS
SaveExploits1References6
Github Security Blog
Github Security Blog
•added 2026/06/02 6:31 p.m.•17 views

OpenMed vulnerable to remote code injection through privacy-filter model loading path

OpenMed before 1.5.2 contains a remote code execution vulnerability in the PII privacy-filter model loading path. The privacy-filter dispatcher used broad substring matching on the user-supplied modelname parameter, allowing a value such as attacker/foo-privacy-filter-bar to route through a path...

9.8CVSS6.4AI score0.01288EPSS
SaveExploits1References6Affected Software1
NVD
NVD
•added 2026/06/02 4:16 p.m.•30 views

CVE-2026-47117

OpenMed before 1.5.2 contains a remote code execution vulnerability in the PII privacy-filter model loading path. The privacy-filter dispatcher used broad substring matching on the user-supplied modelname parameter, allowing a value such as attacker/foo-privacy-filter-bar to route through a path...

9.8CVSS0.01288EPSS
SaveExploits1References5
Rows per page
Query Builder