PREVIEW. This is how it will look in an inbox; it was not sent to you. Each real email carries its own unsubscribe link and Keryn's postal address.

Keryn Headlines

Weekly, week 46 of 2025 · Retrospective · Covers 4 November 2025 12:00 to 11 November 2025 12:00 UTC

Read on the web

Top developments

New LandFall spyware exploited Samsung zero-day via WhatsApp messages

BleepingComputer · Incident · Remediation

CISA orders feds to patch Samsung zero-day used in spyware attacks

BleepingComputer · Advisory · Remediation

GlassWorm malware returns on OpenVSX with 3 new VSCode extensions

BleepingComputer · Analysis · SCA

Critical Cisco UCCX flaw lets attackers run commands as root

BleepingComputer · Advisory · Remediation

A runtime security approach to detecting supply chain attacks

Datadog · Research · SCA

SAST

LLM-Driven Adaptive Source-Sink Identification and False Positive Mitigation for Static Analysis

arXiv · Research · SAST

VulInstruct: Teaching LLMs Root-Cause Reasoning for Vulnerability Detection via Security Specifications

arXiv · Research · Remediation, SAST

Hybrid Fuzzing with LLM-Guided Input Mutation and Semantic Feedback

arXiv · Research · DAST, SAST

SCA

Malicious NuGet packages drop disruptive 'time bombs'

BleepingComputer · Analysis · SCA

Npm Package Targeting GitHub-Owned Repositories Flagged as Red Team Exercise

The Hacker News · Research · SCA

Popular JavaScript library expr-eval vulnerable to RCE flaw

BleepingComputer · Analysis · SCA

Did You Forkget It? Detecting One-Day Vulnerabilities in Open-source ForksWith Global History Analysis

arXiv · Research · Remediation, SCA

DAST

Risk Estimation in Differential Fuzzing via Extreme Value Theory

arXiv · Research · DAST

Agentic SDLC

CVE-2025-21042: Samsung Mobile Devices Out-of-Bounds Write Vulnerability added to CISA's Known Exploited Vulnerabilities catalog

CISA · Advisory · Agentic SDLC

The OpenHands Software Agent SDK: A Composable and Extensible Foundation for Production Agents

arXiv · Research · Agentic SDLC

Remediation

Apple addresses more than 100 vulnerabilities in security updates for iPhones, Macs and iPads

CyberScoop · Analysis · Remediation

From LLMs to Agents: A Comparative Evaluation of LLMs and LLM-based Agents in Security Patch Detection

arXiv · Research · Remediation

Still on Windows 10? Enroll in free ESU before next week’s Patch Tuesday

BleepingComputer · Analysis · Remediation

Hackers Exploiting Triofox Flaw to Install Remote Access Tools via Antivirus Feature

The Hacker News · Analysis · Remediation

Trust

Migrating from Open Policy Agent to Amazon Verified Permissions

AWS Security Blog · Standard / policy · Trust

Models

Class-feature Watermark: A Resilient Black-box Watermark Against Model Extraction Attacks

arXiv · Research · Models

Provable Repair of Deep Neural Network Defects by Preimage Synthesis and Property Refinement

arXiv · Research · Models

JPRO: Automated Multimodal Jailbreaking via Multi-Agent Collaboration Framework

arXiv · Research · Models

Breaking the Stealth-Potency Trade-off in Clean-Image Backdoors with Generative Trigger Optimization

arXiv · Research · Models

AI Security

RAG-targeted Adversarial Attack on LLM-based Threat Detection and Mitigation Framework

arXiv · Research · AI Security, Models

Injecting Falsehoods: Adversarial Man-in-the-Middle Attacks Undermining Factual Recall in LLMs

arXiv · Research · AI Security

Can LLM Infer Risk Information From MCP Server System Logs?

arXiv · Research · AI Security

When AI Meets the Web: Prompt Injection Risks in Third-Party AI Chatbot Plugins

arXiv · Research · AI Security

Advisories in this window

  • 79 npm advisories · 73 critical · 3 high · 1 moderate · 2 low
  • 20 Go advisories · 1 critical · 8 high · 10 moderate · 1 low
  • 11 PyPI advisories · 5 high · 4 moderate · 2 low
  • 4 crates.io advisories · 4 high
  • 3 Maven advisories · 1 high · 1 moderate · 1 low
  • 2 Packagist advisories · 1 high · 1 low
  • 2 Multi-ecosystem advisories · 2 high
  • 2 RubyGems advisories · 2 high

All advisories

Titles, links and short excerpts only; every item links to its source. Selection is automatic; the editor's note is AI-drafted and checked as described. Preview: the unsubscribe link and Keryn's postal address are added to each email when it is sent.