Automatically Attacking Software Reverse Engineering AI Agents

Software tools for reverse engineering executable binary files, such as Ghidra, enable malware analysts to safely conduct robust static analysis without having access to original source code. Coupled with the analytic power of large language models LLM, agentic systems enabled with tools, such as...

CVE-2026-45134

LangSmith Client SDKs provide SDK's for interacting with the LangSmith platform. Prior to LangSmith SDK Python 0.8.0 and JS/TS 0.6.0, the LangSmith SDK's prompt pull methods pullprompt / pullpromptcommit in Python, pullPrompt / pullPromptCommit in JS/TS fetch and deserialize prompt manifests from...

EUVD-2026-32640

LangSmith Client SDKs provide SDK's for interacting with the LangSmith platform. Prior to LangSmith SDK Python 0.8.0 and JS/TS 0.6.0, the LangSmith SDK's prompt pull methods pullprompt / pullpromptcommit in Python, pullPrompt / pullPromptCommit in JS/TS fetch and deserialize prompt manifests from...

CVE-2026-45134

LangSmith CVE-2026-45134 affects LangSmith Client SDKs with prompt-pull methods that fetch/deserialize prompt manifests from LangSmith Hub. The issue allows manifest content to be influenced by external parties when pulling a public prompt (owner/name), because prior SDKs did not distinguish such...

Refusal Before Decoding: Detecting and Exploiting Refusal Signals in Intermediate LLM Activations

In this paper, we investigate whether refusal behavior can be predicted from LLM intermediate activations before decoding using linear probes trained on residual stream activations at each transformer block. We find that refusal is linearly decodable well before the final layer, indicating that...

Cybersecurity AI (CAI) Dataset

We present CAI Dataset, a fourteen-month corpus of cybersecurity LLM trajectories collected through the open-source CAI agent framework, built in response to PentestGPT's finding that expert operator trajectories, not base-model capability, are the bottleneck for cybersecurity LLM performance. CA...

Measuring Real-World Prompt Injection Attacks in LLM-Based Resume Screening

LLMs are vulnerable to prompt injection attacks. However, this vulnerability has been primarily demonstrated conceptually in academic studies or through a few anecdotal case studies. Its prevalence and impact in real-world LLM-based applications are largely unexplored. In this work, we present th...

Malicious code in @catclaw/message-logger-plugin (npm)

--- -= Per source details. Do not edit below this line.=- Source: amazon-inspector cf070f85ba454a799d80e6998ee717f0fc9084513041893a164752162e0b0864 On plugin registration, the log-collector is enabled by default and uploads session JSONL files from /.openclaw/agents//sessions to...

MAL-2026-4782 Malicious code in @catclaw/message-logger-plugin (npm)

--- -= Per source details. Do not edit below this line.=- Source: amazon-inspector cf070f85ba454a799d80e6998ee717f0fc9084513041893a164752162e0b0864 On plugin registration, the log-collector is enabled by default and uploads session JSONL files from /.openclaw/agents//sessions to...

Disentangling Adversarial Prompts: A Semantic-Graph Defense for Robust LLM Security

Large Language Models LLMs are increasingly vulnerable to adversarial prompts that exploit semantic ambiguities to bypass safety mechanisms, resulting in harmful or inappropriate outputs. Such attacks, including jailbreaking and prompt injection, pose significant risks to the integrity and...

Malicious code in cch-agent (PyPI)

--- -= Per source details. Do not edit below this line.=- Source: amazon-inspector cba1bd1e6bb56f0c9816ab482e2ee7cc3a8f04d9e253dd3afa67e4c71b3ae3a2 simpleagent/init.py re-exports ask and chat from simpleagent/client.py. Both entry points ignore caller-supplied configuration and route the caller's...

Reasoning As an Attack Surface: Adaptive Evolutionary CoT Jailbreaks for LLMs

Large Reasoning Models LRMs have demonstrated remarkable capabilities in reasoning and generation tasks and are increasingly deployed in real-world applications. However, their explicit chain-of-thought CoT mechanism introduces new security risks, making them particularly vulnerable to jailbreak...

Google Go 安全漏洞

Google Go is a static, strongly typed, compiled, concurrent programming language with garbage collection features from the American company Google. There is a security vulnerability in Google Go, which stems from keys that accept the ConfirmBeforeUse constraint but never enforce it, resulting in...

Amazon Web Services Kiro CLI 安全漏洞

Amazon Web Services Kiro CLI is a command-line intelligent programming tool provided by Amazon, which supports AI agents, MCP integration, and terminal automation. Versions of the Amazon Web Services Kiro CLI prior to 1.28.0 contained security vulnerabilities. These vulnerabilities stemmed from...

From Preventive to Reactive: How AI Coding Assistants Transform Developers' Security Awareness

AI coding assistants are now central to professional software development, yet their impact on how developers think about and practice security remains poorly understood. While prior work has documented vulnerability rates in AI-generated code, a more fundamental question persists: how do these...

Prompt Overflow: What the Guardrail Inspects Is Not What the Model Infers

Guardrail models a.k.a. safety checkers are widely deployed to screen user inputs before they reach large language models LLMs, serving as a primary defense against prompt injection attacks. Due to strict context constraints, these models handle overlength prompts through truncation or...

MAL-2026-4397 Malicious code in @jemavidev/betteragents-pi (npm)

--- -= Per source details. Do not edit below this line.=- Source: amazon-inspector 3b6e1a3902ad5cc75204b7a6eea3727c6a6c31797d7cfd7a0cd12a64892887bd The package brands itself as an OpenRouter LLM extension and instructs users to obtain a key with the canonical sk-or-v1- prefix from...

Malicious code in @self-evolving-harness/kivo (npm)

--- -= Per source details. Do not edit below this line.=- Source: amazon-inspector ce31b5c287727dabb5479a114843b06b80bbd75db10d74014a00db80b9b321bd The package's LLM pipeline Kivo.ingest → value-gate → OpenAILLMProvider resolves its endpoint via resolveLlmConfig in...

Malicious code in seekcode (npm)

--- -= Per source details. Do not edit below this line.=- Source: amazon-inspector 6f4fe5d868d0434123b1a29a739072fe0e0ec0f2efd1ceda4d2c16ccffecf105 When a user selects the advertised deepseek-cn provider, the package's defaultBaseUrlForProvider function in dist/chunk-6U42R724.js returns...

MAL-2026-4667 Malicious code in seekcode (npm)

--- -= Per source details. Do not edit below this line.=- Source: amazon-inspector 6f4fe5d868d0434123b1a29a739072fe0e0ec0f2efd1ceda4d2c16ccffecf105 When a user selects the advertised deepseek-cn provider, the package's defaultBaseUrlForProvider function in dist/chunk-6U42R724.js returns...