PT-2026-84855

Name of the Vulnerable Software and Affected Versions Hermes version 0.21.0 Description An issue exists where coding agents execute git status or git diff for context gathering before trust prompts are presented to the user. If a repository is delivered as files such as via zip, sync, or USB rath...

CAPTURE: Disentangling Preference Drift from Memory Poisoning in Personalized LLM Agents

Personalized language agents use persistent memory to adapt to users over time, but the same mechanism creates an attack surface. When new information conflicts with stored preferences, an agent must distinguish genuine preference drift from temporary context shifts, ambiguity, or adversarial...

yolobox v0.19.4

██╗ ██╗ ██████╗ ██╗ ██████╗ ██████╗ ██████╗ ██╗ ██╗ ╚██╗ ██╔╝██╔═══██╗██║ ██╔═══██╗██╔══██╗██╔═══██╗╚██╗██╔╝ ╚████╔╝ ██║ ██║██║ ██║ ██║██████╔╝██║ ██║ ╚███╔╝ ╚██╔╝ ██║ ██║██║ ██║ ██║██╔══██╗██║ ██║ ██╔██╗ ██║ ╚██████╔╝███████╗╚██████╔╝██████╔╝╚██████╔╝██╔╝ ██╗ ╚═╝ ╚═════╝ ╚══════╝ ╚═════╝ ╚═════╝...

Reveree: Diagnosing LLM Reverse-Engineering Agents

Reverse engineering RE is critical to security tasks such as malware analysis and vulnerability discovery, and large language model LLM agents are increasingly able to perform it autonomously. Capture-the-flag CTF RE challenges have become the standard proxy for measuring this capability, but...

Alibaba Cloud Linux 3 : 0259: resource-agents security update (Important) (ALINUX3-SA-2026:0259)

The remote Alibaba Cloud Linux 3 host has packages installed that are affected by a vulnerability as referenced in the ALINUX3-SA-2026:0259 advisory. Package updates are available for Alibaba Cloud Linux 3 that fix the following vulnerabilities: CVE-2026-59886: pyasn1 is a generic ASN.1 library f...

Securing Claude Code: The New Compliance API, Local Visibility, and Identity Governance

Claude Code reads files, runs shell commands, invokes MCP tools, and acts through the credentials available on a developer’s machine. Anthropic’s new Compliance API endpoints give security teams their clearest view yet into that activity. They also expose a larger problem: activity logs alone...

A week in security (August 24 – August 30)

Last week on Malwarebytes Labs: Protect your WhatsApp account with new passkey and 2FA upgrades The AI agent swarm that attacked Hugging Face is a warning for the future Flock wants privacy to meet surveillance halfway Fake listings can turn trusted platforms into scam springboards Fake Apple Pay...

ECLIPSE: Self-Evolving Stealthy Prompt Injection Attack against Long-Horizon Agentic Systems

Recently, large language model LLM agents, such as Codex, Claude Code, and OpenClaw, have become capable of planning and executing long-horizon tasks through repeated tool calls. This capability also creates new opportunities for prompt injection. Existing attacks either place the malicious...

onecli v2.0.1

The agent harness built for teams. A pro assistant for companies. Give every employee a secured, sandboxed personal agent. Website · Docs · Discord Quick Start Cloud-hosted: onecli.sh Self-hosted git clone https://github.com/onecli/onecli.git && cd onecli pnpm install pnpm run setup Open...

SIR: Self-Improving Red-Teaming for Compute Use Agents

Computer use agents CUAs are vision-language models that perceive a screen and act on a real operating system through mouse, keyboard, and terminal, and they are increasingly deployed to automate everyday digital tasks. Because they can be exposed to untrusted content while operating, they are...

Moirae: A Multimodal Agent Collaborative Framework for Dynamic Android Malware Detection

The Android ecosystem faces persistent and rapidly evolving malware threats. Existing machine learning detectors are vulnerable to concept drift because they rely on implementation-specific features whose distributions change over time. Large language models LLMs offer strong semantic understandi...

CAITLYN: Can LLM Agents Autonomously Synthesize Defenses against Emerging Injection Attacks?

Prompt injection attacks on Large Language Model LLM agents seek to introduce malicious instructions or content into external text sources retrieved by agents, forcing the underlying LLMs to execute harmful actions outside their benign scope. While current defenses effectively counter known...

LLM-Based Agents for Software and Systems Security: Approaches, Applications, and Assessment

Software and systems security workflows are typically procedural: analysts inspect heterogeneous artifacts, form hypotheses, invoke tools, interpret outputs, and revise plans. Large language model LLM-based agents, which can plan, use tools, retain state, and revise actions across multi-step...

Identifying Agentic Automation with Behavioral Telemetry: Part 2

...

apex v2.4.0-canary.d5c18874

Pensar Apex AI-powered penetration testing using autonomous agents — directly in your terminal. Run blackbox and whitebox pentests that explore, reason, and surface real vulnerabilities. Want to run from the cloud or integrate it with your CI/CD? See Pensar Console. Use Cases Developers Run...

TencentOS Server 3: resource-agents (TSSA-2026:0899)

The version of Tencent Linux installed on the remote TencentOS Server 3 host is prior to tested version. It is, therefore, affected by a vulnerability as referenced in the TSSA-2026:0899 advisory. Package updates are available for TencentOS Server 3 that fix the following vulnerabilities:...

RedEvoAgent: Automatic Red-Teaming Agent with Experience-Driven Skill Evolution

LLM-based agents are increasingly deployed in product-level execution harnesses, where jailbreaks can trigger harmful tool use and persistent state changes, creating greater risks than unsafe text generation alone. Existing automatic red-teaming methods often rely on fixed attacks, while recent...

SPA: Securing Persistent LLM Agents across Queries with Plan-First Information-Flow Control

Large language model LLM agents increasingly operate over untrusted webpages, documents, tools, and persistent states while exercising authority over security-sensitive resources. Existing defenses typically protect either planning or individual tool interactions, but persistent agents face a...

The Guard That Cried Wolf: How Scary Words Make Agent Guardrails Refuse Legitimate Actions

Agent guardrails are checks that approve or refuse each action before an LLM executes it. Sometimes they refuse requests that are genuinely safe. This over-safety blocks deployment when a guardrail refuses an authorized task. Evaluating over-safety is hard: at the boundary an authorized action...

TencentOS Server 3: resource-agents (TSSA-2026:0892)

The version of Tencent Linux installed on the remote TencentOS Server 3 host is prior to tested version. It is, therefore, affected by a vulnerability as referenced in the TSSA-2026:0892 advisory. Package updates are available for TencentOS Server 3 that fix the following vulnerabilities:...