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 5 of 2026 · Retrospective · Covers 20 January 2026 12:00 to 27 January 2026 12:00 UTC

Read on the web

Top developments

CVE-2026-21509: Microsoft Office Security Feature Bypass Vulnerability added to CISA's Known Exploited Vulnerabilities catalog

CISA · Advisory · Remediation

CVE-2026-20045: Cisco Unified Communications Products Code Injection Vulnerability added to CISA's Known Exploited Vulnerabilities catalog

CISA · Advisory · Remediation

SmarterMail auth bypass flaw now exploited to hijack admin accounts

BleepingComputer · Incident · Remediation

CVE-2024-37079: Broadcom VMware vCenter Server Out-of-bounds Write Vulnerability added to CISA's Known Exploited Vulnerabilities catalog

CISA · Advisory · Agentic SDLC, Remediation

Fortinet admins report patched FortiGate firewalls getting hacked

BleepingComputer · Incident · Remediation

SAST

LLM-based Vulnerability Detection at Project Scale: An Empirical Study

arXiv · Research · Agentic SDLC, SAST, SCA

AgenticSCR: An Autonomous Agentic Secure Code Review for Immature Vulnerabilities Detection

arXiv · Research · SAST

MulVul: Retrieval-augmented Multi-Agent Code Vulnerability Detection via Cross-Model Prompt Evolution

arXiv · Research · SAST

An Effective and Cost-Efficient Agentic Framework for Ethereum Smart Contract Auditing

arXiv · Research · DAST, SAST, Trust

SCA

Attention-Enhanced Graph Filtering for False Data Injection Attack Detection and Localization

arXiv · Research · SCA

Some ChatGPT browser extensions are stealing your data

CyberScoop · Analysis · SCA

Hackers can bypass npm’s Shai-Hulud defenses via Git dependencies

BleepingComputer · Analysis · SCA

Multi-Agent End-to-End Vulnerability Management for Mitigating Recurring Vulnerabilities

arXiv · Research · Remediation, SAST, SCA

DAST

Why has Microsoft been routing example.com traffic to a company in Japan?

Ars Technica: Security · Analysis · DAST

Data-driven Test Generation for Fuzzing AI Compiler

arXiv · Research · DAST

FirmReBugger: A Benchmark Framework for Monolithic Firmware Fuzzers

arXiv · Research · DAST

SmartOracle -- An Agentic Approach to Mitigate Noise in Differential Oracles

arXiv · Research · DAST

Agentic SDLC

Malicious AI extensions on VSCode Marketplace steal developer data

BleepingComputer · Analysis · Agentic SDLC, AI Security

Understanding Dominant Themes in Reviewing Agentic AI-authored Code

arXiv · Research · Agentic SDLC

Prompt Injection Attacks on Agentic Coding Assistants: A Systematic Analysis of Vulnerabilities in Skills, Tools, and Protocol Ecosystems

arXiv · Research · Agentic SDLC, AI Security

PatchIsland: Orchestration of LLM Agents for Continuous Vulnerability Repair

arXiv · Research · Agentic SDLC, Remediation

Remediation

OpenSSL January 2026 Security Update: CMS and PKCS#12 Buffer Overflows

Datadog · Analysis · Remediation

AI-supported vulnerability triage with the GitHub Security Lab Taskflow Agent

GitHub · Analysis · Remediation

VulnResolver: A Hybrid Agent Framework for LLM-Based Automated Vulnerability Issue Resolution

arXiv · Research · Remediation

Power agentic workflows in your terminal with GitHub Copilot CLI

GitHub · Analysis · Agentic SDLC, Remediation

Trust

MultiChain Blockchain Data Provenance for Deterministic Stream Processing with Kafka Streams: A Weather Data Case Study

arXiv · Research · Trust

Faramesh: A Protocol-Agnostic Execution Control Plane for Autonomous Agent Systems

arXiv · Research · Trust

PAL*M: Property Attestation for Large Generative Models

arXiv · Research · Trust

Preserving Open Source Sustainability While Advancing Cybersecurity Compliance

OpenSSF · Analysis · SCA, Trust

Models

LLM-VA: Resolving the Jailbreak-Overrefusal Trade-off via Vector Alignment

arXiv · Research · Models

From Internal Diagnosis to External Auditing: A VLM-Driven Paradigm for Data-Free Online Backdoor Defense

arXiv · Research · Models

Thought-Transfer: Indirect Targeted Poisoning Attacks on Chain-of-Thought Reasoning Models

arXiv · Research · Models

Comparison requires valid measurement: Rethinking attack success rate comparisons in AI red teaming

arXiv · Research · Models

AI Security

From runtime risk to real‑time defense: Securing AI agents

Microsoft · Analysis · AI Security

Prompt Injection Evaluations: Refusal Boundary Instability and Artifact-Dependent Compliance in GPT-4-Series Models

arXiv · Research · AI Security, Models

Breaking the Protocol: Security Analysis of the Model Context Protocol Specification and Prompt Injection Vulnerabilities in Tool-Integrated LLM Agents

arXiv · Research · AI Security

Physical Prompt Injection Attacks on Large Vision-Language Models

arXiv · Research · AI Security

Advisories in this window

  • 176 npm advisories · 149 critical · 16 high · 8 moderate · 3 low
  • 42 Go advisories · 3 critical · 14 high · 19 moderate · 6 low
  • 23 PyPI advisories · 11 high · 9 moderate · 3 low
  • 22 Maven advisories · 3 critical · 5 high · 9 moderate · 5 low
  • 12 Packagist advisories · 1 critical · 2 high · 8 moderate · 1 low
  • 5 crates.io advisories · 2 high · 2 moderate · 1 low
  • 5 NuGet advisories · 4 moderate · 1 low
  • 2 Swift advisories · 1 moderate · 1 low
  • 1 Multi-ecosystem advisory · 1 critical
  • 1 RubyGems advisory · 1 moderate

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.