Design Principles for the Construction of a Benchmark Evaluating Security Operation Capabilities of Multi-Agent AI Systems

As Large Language Models LLMs and multi-agent AI systems are demonstrating increasing potential in cybersecurity operations, organizations, policymakers, model providers, and researchers in the AI and cybersecurity communities are interested in quantifying the capabilities of such AI systems to...

EUVD-2026-17026

OpenClaw before 2026.3.11 contains an authorization bypass vulnerability in the gateway agent RPC that allows authenticated operators with operator.write permission to override workspace boundaries by supplying attacker-controlled spawnedBy and workspaceDir values. Remote operators can escape the...

CVE-2026-33573

OpenClaw (prior to 2026.3.11) suffers an authorization bypass in the gateway agent RPC. Authenticated operators with operator.write permission can override workspace boundaries by supplying attacker-controlled spawnedBy and workspaceDir values, enabling escape from the configured workspace and ex...

CVE-2026-33573 OpenClaw < 2026.3.11 - Workspace Boundary Bypass via Agent RPC Parameters

OpenClaw before 2026.3.11 contains an authorization bypass vulnerability in the gateway agent RPC that allows authenticated operators with operator.write permission to override workspace boundaries by supplying attacker-controlled spawnedBy and workspaceDir values. Remote operators can escape the...

CVE-2026-33573 OpenClaw < 2026.3.11 - Workspace Boundary Bypass via Agent RPC Parameters

OpenClaw before 2026.3.11 contains an authorization bypass vulnerability in the gateway agent RPC that allows authenticated operators with operator.write permission to override workspace boundaries by supplying attacker-controlled spawnedBy and workspaceDir values. Remote operators can escape the...

CVE-2026-32919 OpenClaw < 2026.3.11 - Unauthorized Session Reset via agent Slash Commands

OpenClaw before 2026.3.11 contains an authorization bypass vulnerability allowing write-scoped callers to reach admin-only session reset logic. Attackers with operator.write scope can issue agent requests containing /new or /reset slash commands to reset targeted conversation state without holdin...

Deserialization Of Untrusted Data

io.opentelemetry.javaagent:opentelemetry-javaagent is vulnerable to Deserialization Of Untrusted Data. The vulnerability is due to lack of serialization filtering in the RMI instrumentation endpoint, which allows an attacker with network access to send malicious serialized data and execute...

Deserialization Of Untrusted Data

com.datadoghq:dd-java-agent is vulnerable to Deserialization of Untrusted Data. The vulnerability is due to unsafe deserialization of incoming data by a custom RMI endpoint without applying serialization filters, which allows a network-based attacker to exploit a gadget chain and potentially...

CVE-2026-33728

dd-trace-java is a Datadog APM client for Java. In versions of dd-trace-java 0.40.0 through prior to 1.60.2, the RMI instrumentation registered a custom endpoint that deserialized incoming data without applying serialization filters. On JDK version 16 and earlier, an attacker with network access ...

SUSE CVE-2026-24516

A command injection vulnerability exists in DigitalOcean Droplet Agent through 1.3.2. The troubleshooting actioner component internal/troubleshooting/actioner/actioner.go processes metadata from the metadata service endpoint and executes commands specified in the TroubleshootingAgent.Requesting...

SUSE CVE-2026-25783

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...

SUSE CVE-2026-27859

A mail message containing excessive amount of RFC 2231 MIME parameters causes LMTP to use too much CPU. A suitably formatted mail message causes mail delivery process to consume large amounts of CPU time. Use MTA capabilities to limit RFC 2231 MIME parameters in mail messages, or upgrade to fixed...

OPENSUSE-SU-2026:10453-1 libpolkit-agent-1-0-127-3.1 on GA media

These are all security issues fixed in the libpolkit-agent-1-0-127-3.1 package on the GA media of openSUSE Tumbleweed...

SafeClaw-R: Towards Safe and Secure Multi-Agent Personal Assistants

LLM-based multi-agent systems MASs are transforming personal productivity by autonomously executing complex, cross-platform tasks. Frameworks such as OpenClaw demonstrate the potential of locally deployed agents integrated with personal data and services, but this autonomy introduces significant...

Red-MIRROR: Agentic LLM-Based Autonomous Penetration Testing with Reflective Verification and Knowledge-Augmented Interaction

Web applications remain the dominant attack surface in cybersecurity, where vulnerabilities such as SQL injection, XSS, and business logic flaws continue to cause significant data breaches. While penetration testing is effective for identifying these weaknesses, traditional manual approaches are...

amazon-cloudwatch-agent-1.300064.0-2.1 on GA media (moderate)

amazon-cloudwatch-agent-1.300064.0-2.1 on GA media Announcement ID: openSUSE-SU-2026:10420-1 Rating: moderate Cross-References: CVE-2026-33186 CVSS scores: CVE-2026-33186 SUSE : 8.1 CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:N CVE-2026-33186 SUSE : 8.6...

OpenClaw: BlueBubbles Group Reactions Bypass requireMention and Still Enqueue Agent-Visible System Events

Summary BlueBubbles Group Reactions Bypass requireMention and Still Enqueue Agent-Visible System Events Affected Packages / Versions - Package: openclaw - Affected versions: = 2026.3.24 - First patched version: 2026.3.25 - Latest published npm version at verification time: 2026.3.24 Details...

GHSA-MW7W-G3MG-XQM7 OpenClaw: BlueBubbles Group Reactions Bypass requireMention and Still Enqueue Agent-Visible System Events

Summary BlueBubbles Group Reactions Bypass requireMention and Still Enqueue Agent-Visible System Events Affected Packages / Versions - Package: openclaw - Affected versions: = 2026.3.24 - First patched version: 2026.3.25 - Latest published npm version at verification time: 2026.3.24 Details...

CVE-2026-31945

LibreChat is a ChatGPT clone with additional features. Versions 0.8.2-rc2 through 0.8.2 are vulnerable to a server-side request forgery SSRF attack when using agent actions or MCP. Although a previous SSRF vulnerability...

CVE-2026-31945

LibreChat is a ChatGPT clone with additional features. Versions 0.8.2-rc2 through 0.8.2 are vulnerable to a server-side request forgery SSRF attack when using agent actions or MCP. Although a previous SSRF vulnerability...