PT-2026-48898

Name of the Vulnerable Software and Affected Versions ChromaDB versions 0.4.17 through 0.4.16 Description An authenticated attacker with the UPDATE COLLECTION permission can execute arbitrary code on the server. This occurs by sending a malicious model repository and setting the trust remote code...

PT-2026-49031

Name of the Vulnerable Software and Affected Versions OpenClaw versions prior to 2026.5.2 Description An issue in message.action forwarding allows model-controlled metadata to forward action payloads containing Gateway credentials to attacker-supplied loopback URLs. Remote attackers can intercept...

Evaluating LLMs for Obfuscation Detection and Classification in Android Apps

Android applications apps developers increasingly rely on code obfuscation techniques to hinder reverse engineering and protect intellectual property. However, obfuscation also reduces the effectiveness of static analysis and vulnerability detection tools, creating challenges for Android security...

Security Engineering of OpenClaw: Analyzing Attack Surface Expansion and Trust-Boundary Violations

Agentic large language model LLM systems can now execute actions, not only produce text. When model outputs trigger privileged operations such as shell commands, browser automation, or external tool calls, the security problem shifts from alignment alone to system configuration and structural...

MINI-G255-XXR7-QQJW

Bulletin has no description...

MINI-2C48-JXPP-W45G

Bulletin has no description...

MINI-H9XV-H96J-848M

Bulletin has no description...

MINI-RJP8-XC83-R6HQ

Bulletin has no description...

MINI-5X68-866F-V52F

Bulletin has no description...

MINI-878G-P8RJ-CQ5M

Bulletin has no description...

MINI-RCGX-HGC5-45FX

Bulletin has no description...

keras-eval (>=0.0.4, <=0.0.8), keras-model-specs (>=0.0.17, <=0.0.24) +3 more potentially affected by CVE-2026-11816 via keras (2.1.4)

keras PYPI version 2.1.4 is affected by a known vulnerability. The following packages have a transitive dependency on keras and may be impacted: - keras-eval =0.0.4, =0.0.17, =0.0.21, =0.1.2, =0.5.2 Source cves: CVE-2026-11816 Source advisory: OSV:GHSA-HQP4-2352-XF5R...

gesture-classifying-model (>=0.1.2, <=0.1.3) potentially affected by CVE-2026-11816 via keras (1.2.2)

keras PYPI version 1.2.2 is affected by a known vulnerability. The following packages have a transitive dependency on keras and may be impacted: - gesture-classifying-model =0.1.2, =0.1.3 Source cves: CVE-2026-11816 Source advisory: OSV:GHSA-HQP4-2352-XF5R...

gesture-classifying-model (>=0.1.2, <=0.1.3) potentially affected by CVE-2026-11816 via keras (1.2.2)

keras PYPI version 1.2.2 is affected by a known vulnerability. The following packages have a transitive dependency on keras and may be impacted: - gesture-classifying-model =0.1.2, =0.1.3 Source cves: CVE-2026-11816 Source advisory: OSV:PYSEC-2026-2324...

Insufficiently Protected Credentials

Overview meta-ads-mcp is a Model Context Protocol MCP server for Meta Ads - Use Remote MCP at pipeboard.co for easiest setup Affected versions of this package are vulnerable to Insufficiently Protected Credentials via the AuthInjectionMiddleware.dispatch process. An attacker can gain unauthorized...

MINI-MC42-68PH-R2G5

Bulletin has no description...

MINI-77M3-X2C2-CX62

Bulletin has no description...

SUSE CVE-2026-47165

ImageMagick is free and open-source software used for editing and manipulating digital images. Prior to versions 6.9.13-48 and 7.1.2-23, the distributed pixel cache was originally designed to operate without a challenge-response authentication model. This has been changed in versions 6.9.13-48 an...

Arbitrary File Write

open-webui/open-webui is vulnerable to an arbitrary file write. The vulnerability is due to improper handling of file paths in the downloadmodel endpoint on Windows, which allows an attacker to manipulate file paths and write files to arbitrary locations on the server...

CVE-2026-46432

LMDeploy is a toolkit for compressing, deploying, and serving large language models. In versions 0.12.3 and prior, LMDeploy is vulnerable to arbitrary code execution through hardcoded "trustremotecode=True" in multiple HuggingFace model-loading call sites. At time of publication, there are no...