Benchmarking-Agent-Architectures

Benchmarking Agent Architectures for LLM-Based Exploit Gener...

From Incomplete Architecture to Quantified Risk: Multimodal LLM-Driven Security Assessment for Cyber-Physical Systems

Cyber-physical systems often contend with incomplete architectural documentation or outdated information resulting from legacy technologies, knowledge management gaps, and the complexity of integrating diverse subsystems over extended operational lifecycles. This architectural incompleteness...

Text Generation Web UI 路径遍历漏洞

Text Generation Web UI is a local AI UI interface developed by oobabooga’s individual developers. Versions of Text Generation Web UI prior to 4.3 contained a path traversal vulnerability. This vulnerability stemmed from an unauthenticated path traversal vulnerability in the loadprompt function,...

Swiss-Bench 003: Evaluating LLM Reliability and Adversarial Security for Swiss Regulatory Contexts

The deployment of large language models LLMs in Swiss financial and regulatory contexts demands empirical evidence of both production reliability and adversarial security, dimensions not jointly operationalized in existing Swiss-focused evaluation frameworks. This paper introduces Swiss-Bench 003...

PT-2026-30860

text-generation-webui is an open-source web interface for running Large Language Models. Prior to 4.3, an unauthenticated path traversal vulnerability in load prompt allows reading any .txt file on the server filesystem. The file content is returned verbatim in the API response. This vulnerabilit...

EUVD-2026-19440

Anthropic Claude Code CLI and Claude Agent SDK contain an OS command injection vulnerability in the prompt editor invocation utility that allows attackers to execute arbitrary commands by crafting malicious file paths. Attackers can inject shell metacharacters such as $ or backtick expressions in...

CVE-2026-35021

Rejected reason: This CVE ID has been rejected by its CVE Numbering Authority CNA. It was determined that the affected code path cannot be triggered through normal usage of Claude Code...

CVE-2026-35021

This CVE ID has been rejected by its CVE Numbering Authority CNA. It was determined that the affected code path cannot be triggered through normal usage of Claude Code...

CVE-2026-35021

The CVE-2026-35021 entry is rejected by the CNA and does not represent an active vulnerability.

CVE-2026-35021

...

CVE-2026-35021 Anthropic Claude Code & Agent SDK OS Command Injection via promptEditor.ts

Anthropic Claude Code CLI and Claude Agent SDK contain an OS command injection vulnerability in the prompt editor invocation utility that allows attackers to execute arbitrary commands by crafting malicious file paths. Attackers can inject shell metacharacters such as $ or backtick expressions in...

CVE-2026-35021

...

Mapping the Exploitation Surface: A 10,000-Trial Taxonomy of What Makes LLM Agents Exploit Vulnerabilities

LLM agents with tool access can discover and exploit security vulnerabilities. This is known. What is not known is which features of a system prompt trigger this behaviour, and which do not. We present a systematic taxonomy based on approximately 10,000 trials across seven models, 37 prompt...

PT-2026-30707

Name of the Vulnerable Software and Affected Versions Anthropic Claude Code CLI and Claude Agent SDK affected versions not specified Description The Anthropic Claude Code CLI and Claude Agent SDK contain an OS command injection issue in the prompt editor invocation utility. Attackers can execute...

Claude Code CLI和Claude Agent SDK 操作系统命令注入漏洞

Claude Code CLI and Claude Agent SDK are both open-source products developed by Anthropic. Claude Code CLI is a command-line AI coding assistant tool. Claude Agent SDK is a developer toolkit for AI coding assistants. Both Claude Code CLI and Claude Agent SDK have operating system command injectio...

LLM-Enabled Open-Source Systems in the Wild: An Empirical Study of Vulnerabilities in GitHub Security Advisories

Large language models LLMs are increasingly embedded in open-source software OSS ecosystems, creating complex interactions among natural language prompts, probabilistic model outputs, and execution-capable components. However, it remains unclear whether traditional vulnerability disclosure...

PT-2026-30458

Name of the Vulnerable Software and Affected Versions Dompdf versions prior to 3.16 Description Dompdf is subject to a File Existence Oracle attack via the manipulation of the CSS @font-face directive. An attacker can provide malicious HTML that repeatedly references local files using the file://...

@mobilenext/mobile-mcp: Arbitrary Android Intent Execution via mobile_open_url

Summary The mobileopenurl tool in mobile-mcp passes user-supplied URLs directly to Android's intent system without any scheme validation, allowing execution of arbitrary Android intents, including USSD codes, phone calls, SMS messages, and content provider access. Details The vulnerable code pass...

CVE-2026-34779

Electron is a framework for writing cross-platform desktop applications using JavaScript, HTML and CSS. Prior to versions 38.8.6, 39.8.1, 40.8.0, and 41.0.0-beta.8, on macOS, app.moveToApplicationsFolder used an AppleScript fallback path that did not properly handle certain characters in the...

CVE-2026-34779 Electron: AppleScript injection in app.moveToApplicationsFolder on macOS

Electron is a framework for writing cross-platform desktop applications using JavaScript, HTML and CSS. Prior to versions 38.8.6, 39.8.1, 40.8.0, and 41.0.0-beta.8, on macOS, app.moveToApplicationsFolder used an AppleScript fallback path that did not properly handle certain characters in the...