1940 matches found
EUVD-2026-13974
OpenClaw versions prior to 2026.2.22 reuse gateway.auth.token as a fallback hash secret for owner-ID prompt obfuscation when commands.ownerDisplay is set to hash and commands.ownerDisplaySecret is unset, creating dual-use of authentication secrets across security domains. Attackers with access to...
CVE-2026-32897 OpenClaw < 2026.2.22 - Authentication Token Reuse in Owner ID Prompt Hashing Fallback
OpenClaw versions prior to 2026.2.22 reuse gateway.auth.token as a fallback hash secret for owner-ID prompt obfuscation when commands.ownerDisplay is set to hash and commands.ownerDisplaySecret is unset, creating dual-use of authentication secrets across security domains. Attackers with access to...
CVE-2026-32897
OpenClaw versions prior to 2026.2.22 reuse gateway.auth.token as a fallback hash secret for owner-ID prompt obfuscation when commands.ownerDisplay is set to hash and commands.ownerDisplaySecret is unset. This allows attackers with access to system prompts sent to third-party model providers to de...
CVE-2026-32897 OpenClaw < 2026.2.22 - Authentication Token Reuse in Owner ID Prompt Hashing Fallback
OpenClaw versions prior to 2026.2.22 reuse gateway.auth.token as a fallback hash secret for owner-ID prompt obfuscation when commands.ownerDisplay is set to hash and commands.ownerDisplaySecret is unset, creating dual-use of authentication secrets across security domains. Attackers with access to...
OpenClaw 安全漏洞
OpenClaw is an intelligent artificial assistant open-sourced by OpenClaw. OpenClaw has a security vulnerability that can be exploited by an attacker to cause bypassing of interactive approval prompts...
PT-2026-26747
OpenClaw versions prior to 2026.2.23 contain an authorization bypass vulnerability in the ACP client that auto-approves tool calls based on untrusted toolCall.kind metadata and permissive name heuristics. Attackers can bypass interactive approval prompts for read-class operations by spoofing tool...
T-MAP: Red-Teaming LLM Agents with Trajectory-Aware Evolutionary Search
While prior red-teaming efforts have focused on eliciting harmful text outputs from large language models LLMs, such approaches fail to capture agent-specific vulnerabilities that emerge through multi-step tool execution, particularly in rapidly growing ecosystems such as the Model Context Protoc...
OpenClaw 安全漏洞
OpenClaw is an open-source intelligent artificial assistant developed by OpenClaw. Versions of OpenClaw prior to 2026.2.24 contained security vulnerabilities. These vulnerabilities stemmed from a permission bypass issue in the system.run allowlist mode, which allowed attackers to execute commands...
Security Assessment and Mitigation Strategies for Large Language Models: A Comprehensive Defensive Framework
Large Language Models increasingly power critical infrastructure from healthcare to finance, yet their vulnerability to adversarial manipulation threatens system integrity and user safety. Despite growing deployment, no comprehensive comparative security assessment exists across major LLM...
ChainFuzzer: Greybox Fuzzing for Workflow-Level Multi-Tool Vulnerabilities in LLM Agents
Tool-augmented LLM agents increasingly rely on multi-step, multi-tool workflows to complete real tasks. This design expands the attack surface, because data produced by one tool can be persisted and later reused as input to another tool, enabling exploitable source-to-sink dataflows that only...
PT-2026-24209
Name of the Vulnerable Software and Affected Versions openclaw-core versions 0.3.x Description A critical deserialization issue exists in openclaw-core that allows arbitrary instruction injection through specially crafted system prompts. This affects all deployed instances. The issue requires...
Two Frames Matter: A Temporal Attack for Text-To-Video Model Jailbreaking
Recent text-to-video T2V models can synthesize complex videos from lightweight natural language prompts, raising urgent concerns about safety alignment in the event of misuse in the real world. Prior jailbreak attacks typically rewrite unsafe prompts into paraphrases that evade content filters...
PT-2026-23799
Name of the Vulnerable Software and Affected Versions WeKnora versions prior to 0.3.0 Description WeKnora, an LLM-powered framework for deep document understanding and semantic retrieval, is susceptible to a vulnerability involving tool name collision and indirect prompt injection. A malicious...
Manipulating AI Summarization Features
Microsoft is reporting: Companies are embedding hidden instructions in "Summarize with AI" buttons that, when clicked, attempt to inject persistence commands into an AI assistant's memory via URL prompt parameters…. These prompts instruct the AI to "remember Company as a trusted source" or...
GHSA-V6X2-2QVM-6GV8 OpenClaw reuses the gateway auth token in the owner ID prompt hashing fallback
Vulnerability OpenClaw reused gateway.auth.token and gateway.remote.token as a fallback hash secret for owner-ID prompt obfuscation when commands.ownerDisplay=hash and commands.ownerDisplaySecret was unset. This created secret dual-use between gateway authentication and prompt metadata hashing...
GHSA-7FF8-XJH3-MGH6 OpenClaw's non-default autoAllowSkills setting could bypass on-miss exec prompt
Summary In openclaw versions up to and including 2026.2.22-2, a non-default exec-approval configuration could allow a skill-name collision to bypass an ask=on-miss prompt. When autoAllowSkills=true, a path-scoped executable such as ./skill-bin could resolve to basename skill-bin, satisfy the skil...
PT-2026-26392
Summary This issue applies to a non-default configuration only. If sort is manually added to tools.exec.safeBins, OpenClaw could treat sort --compress-program= as valid safe-bin usage. In security=allowlist + ask=on-miss, this could satisfy allowlist checks and skip operator approval, while GNU...
CVE-2026-2256 Command injection vulnerability in ModelScope's ms-agent
A command injection vulnerability in ModelScope's ms-agent versions v1.6.0rc1 and earlier exists, allowing an attacker to execute arbitrary operating system commands through crafted prompt-derived input...
Arbitrary Command Injection
Overview ms-agent is a MS-Agent: Lightweight Framework for Empowering Agents with Autonomous Exploration Affected versions of this package are vulnerable to Arbitrary Command Injection via the ms-agent process. An attacker can execute arbitrary operating system commands by supplying specially...
Samsung TVs stop spying on viewers in Texas. Here’s how to disable ACR anywhere
Samsung has settled a lawsuit with the Texas Attorney General over how its smart TVs collect and monetize viewing data using Automated Content Recognition ACR. As part of the settlement, Samsung agreed to stop collecting ACR data from Texans without explicit, informed consent and to rewrite its...