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 42 of 2026 · Retrospective · Covers 6 October 2026 11:00 to 13 October 2026 11:00 UTC

Read on the web

Top developments

Tensorlake npm Package Compromised: A Worm With a Hostage Token That Wipes Your Machine If You Revoke It

StepSecurity · Incident · SCA, Trust

Citrix warns admins to patch new NetScaler RCE flaw immediately

BleepingComputer · Advisory · Remediation

Attackers Exploit CVSS 10.0 Bricksforge WordPress Flaw Enabling Unauthenticated RCE

The Hacker News · Analysis · Remediation

ARTEX AI, Claude agents used in cyberattacks on South Korean banks

BleepingComputer · Analysis · DAST

SonicWall warns of max severity SSRF flaw in SMA1000 gateways

BleepingComputer · Analysis · Remediation

SAST

LLM-based Code Auditing by Inferring and Evaluating Developer Assumptions

arXiv · Research · SAST

EvolvAttack: Backdooring self Evolving VLM Agents

arXiv · Research · Models, SAST

SecReason: Vulnerability Detection and In-Stream Self-Repair for LLM Code Generation

arXiv · Research · SAST

ObliVul: Alert-Conditioned Safety Obligation Modeling and Bidirectional Counterfactual Validation for Code Vulnerability Detection

arXiv · Research · SAST

SCA

VulnBump: A Benchmark for Fixing Vulnerable Dependencies in Real Projects While Keeping the Build and Tests Green

arXiv · Research · Agentic SDLC, Remediation, SCA

And another Tool!

SPDX · Analysis · AI-BOM, SCA

Auditing the auditor: Unmasking Apex Accelerator

Elastic · Analysis · Agentic SDLC, AI Security, SCA

Credential-Stealing GitHub Actions Workflows Planted in Tens of Thousands of Repositories

The Hacker News · Research · SCA

DAST

Aikido Security is now CREST accredited

Aikido Security · Analysis · DAST

What Is Agentic Pentesting? What It Proves, and Where It Stops.

The Hacker News · Guide · DAST

Agentic SDLC

How one bug bounty researcher chooses the features they investigate

GitHub · Research · Agentic SDLC, Remediation

CyberXero Uses Claude Code in Campaign That Stole Over 628,000 Ukrainian Records

The Hacker News · Incident · Agentic SDLC

Last Secure Code Benchmark: Is the Code You Vibe Secure?

arXiv · Research · Agentic SDLC

Names Are Not Objects: Object-Bound Confinement for Compromised AI Coding Agents

arXiv · Research · Agentic SDLC, AI Security

AI-BOM

Vulnerable Models in the AI Supply Chain: An Empirical Study of Adoption, Propagation, and Usage Patterns

arXiv · Research · AI-BOM

Remediation

An Industry Perspective on Software Vulnerability Management: A Systematic Review of Approaches, Tools, Practices, and Challenges

arXiv · Research · Remediation

CWE-bench: Refocusing Agentic Security from Exploitation to Defense

arXiv · Research · AI Security, Remediation

GitLab warns of max severity code execution vulnerability

BleepingComputer · Analysis · Remediation

Natural Code Transformation Confuses Agents in Vulnerability Analysis

arXiv · Research · AI Security, Remediation

Trust

You Won’t Hear About These, Even In Myths (Atlassian Jira, Confluence (and more) Pre-Auth Arbitrary File Read CVE-2026-21589)

watchTowr · Analysis · Trust

Computational Reflective Equilibrium for Agentic AI: A Rawlsian Alignment Scaffold for Governance

arXiv · Research · Trust

ANASSA: From High-Level Framework to Low-Level Architecture for Spatial Agentic AI

arXiv · Research · Trust

CYBERFORT: A Compliance-Chain Platform Operationalising the Cyber Resilience Act for SMEs

arXiv · Research · Trust

Models

TRAP: Tree-Guided Reinforcement-Augmented Probing for Automated LLM Jailbreaking

arXiv · Research · Models

CircuitBreak: Pruning a Sparse Safety Circuit of Neurons

arXiv · Research · Models

ATTCK-R: Knowledge-Aware Long-Horizon Jailbreaking for Protected LLMs

arXiv · Research · Models

Adversarial Robustness of Camera-Radar Sensor Fusion for Autonomous Driving

arXiv · Research · Models

AI Security

ProtoArena: Can AI Agents Discover and Exploit Real-World Cross-Component Protocol Vulnerabilities?

arXiv · Research · AI Security

BrowseSafe++: Benchmarking Prompt Injection Detection Across Web Contexts, Languages, and Browser Views

arXiv · Research · AI Security

StableJudge: A Lightweight Adversarial Defense Framework for LLM-as-a-Judge

arXiv · Research · AI Security

Only What You've Signed up For: Secure Workflows with Constrained Diffusion Agents

arXiv · Research · AI Security

Advisories in this window

  • 404 npm advisories · 279 critical · 61 high · 54 moderate · 10 low
  • 97 PyPI advisories · 17 critical · 36 high · 40 moderate · 4 low
  • 75 Go advisories · 4 critical · 37 high · 33 moderate · 1 low
  • 26 RubyGems advisories · 24 critical · 2 moderate
  • 18 Maven advisories · 2 critical · 5 high · 9 moderate · 2 low
  • 12 NuGet advisories · 1 critical · 7 high · 3 moderate · 1 low
  • 11 Packagist advisories · 1 critical · 1 high · 8 moderate · 1 low
  • 1 crates.io advisory · 1 critical

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.