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 26 of 2026 · Retrospective · Covers 16 June 2026 11:00 to 23 June 2026 11:00 UTC

Read on the web

Top developments

Mastra npm Supply Chain Attack: 140+ Packages Backdoored via easy-day-js Typosquat

StepSecurity · Incident · SCA

CISA: Splunk Enterprise flaw actively exploited, patch by Sunday

BleepingComputer · Incident · Remediation

CISA orders feds to patch max severity Joomla plugin flaw by Friday

BleepingComputer · Advisory · Remediation

Microsoft working on Defender patch for RoguePlanet zero-day

BleepingComputer · Advisory · Remediation

Apple fixes Beats Studio Buds flaw that let hackers spy on conversations

BleepingComputer · Analysis · Remediation

SAST

Revelio: Cost-Efficient Agentic Memory Safety Vulnerability Detection For Repository-Scale Codebases

arXiv · Research · SAST

Evaluating LLMs for Real-World Web Vulnerability Detection

arXiv · Research · Agentic SDLC, SAST

"What Happens Locally, Leaks Globally": Detecting Privacy Leakage Risks in MCP Servers

arXiv · Research · AI Security, SAST

Calibration Without Comprehension: Diagnosing the Limits of Fine-Tuning LLMs for Vulnerability Detection in Systems Software

arXiv · Research · Remediation, SAST

SCA

Detecting the Klue supply chain attack in Salesforce instances

Datadog · Incident · SCA

Microsoft links Mastra AI supply chain attack to North Korean hackers

BleepingComputer · Incident · SCA

ShapedPlugin update flow hacked to infect WordPress sites

BleepingComputer · Incident · SCA

400+ AUR Packages Hijacked: What the “Atomic Arch” Campaign Means for Supply-Chain Security

StepSecurity · Incident · SCA

Agentic SDLC

Malicious JetBrains Marketplace plugins steal AI API keys from developers

BleepingComputer · Analysis · Agentic SDLC

Critical Copilot vulnerability allowed hackers to steal 2FA code from users

Ars Technica: Security · Research · Agentic SDLC, Remediation

The Hitchhiker's Guide to Agentic AI: From Foundations to Systems

arXiv · Research · Agentic SDLC

Understanding the (In)Security of Vibe-Coded Applications

arXiv · Research · Agentic SDLC

AI-BOM

AgentRiskBOM: A Risk-Scoping Security Bill of Materials for Agentic AI Systems

arXiv · Research · AI-BOM, AI Security, SCA

Execution-bound advisory automation for agentic AI: a reproducible AIBOM-driven CSAF-VEX framework

arXiv · Research · AI-BOM, Trust

Remediation

Hackers exploit info disclosure bug in Gravity SMTP WordPress plugin

BleepingComputer · Analysis · Remediation

OpenAI Expands Daybreak With GPT-5.5-Cyber to Help Defenders Patch Security Flaws

The Hacker News · Advisory · Remediation

From vulnerability report to CVE draft in minutes: how Elastic automated security advisories with AI

Elastic · Analysis · AI Security, Remediation

VeriPort: Automated and Verified Patch Backporting at Scale

arXiv · Research · Remediation, SCA

Trust

AI Security Priorities: A Field-Wide Agenda

arXiv · Research · Trust

The Human Utility Factor: A Computable Welfare Metric That Reframes AI Governance as a Constrained Optimisation Problem

arXiv · Research · Trust

The New Security Control Point: Governing AI Agents Inside the Execution Loop

Snyk · Research · AI Security, Trust

LLM Unlearning for Cyber Defense: A Survey on Methods, Challenges, and Emerging Threats

arXiv · Research · Models, Trust

Models

PixJail: Self-Evolving Paper-to-Pipeline Reproduction for Text-to-Image Jailbreak Evaluation

arXiv · Research · Models

TROPT: An Open Framework for Unifying and Advancing Discrete Text Optimization

arXiv · Research · Models

TooBad: Backdoor Diffusion Models with Ultra-Low Poison Rate and Imperceptible Trigger

arXiv · Research · Models

Rethinking Molecular Graph Backdoors under Chemistry-aware Admission

arXiv · Research · Models

AI Security

Police cleans nearly 15,000 SocGholish-infected sites tied to Evil Corp

BleepingComputer · Analysis · AI Security

AutoJack: How a single page can RCE the host running your AI agent

Microsoft · Analysis · AI Security

Poisoned Playbooks: Demystifying Knowledge Poisoning Effects on AI Security Agents

arXiv · Research · AI Security

Securing LLM-Agent Long-Term Memory Against Poisoning: Non-Malleable, Origin-Bound Authority with Machine-Checked Guarantees

arXiv · Research · AI Security

Advisories in this window

  • 328 npm advisories · 182 critical · 65 high · 66 moderate · 15 low
  • 151 PyPI advisories · 26 critical · 67 high · 47 moderate · 11 low
  • 91 Go advisories · 5 critical · 38 high · 39 moderate · 9 low
  • 51 Packagist advisories · 4 critical · 16 high · 28 moderate · 3 low
  • 38 Maven advisories · 6 critical · 14 high · 16 moderate · 2 low
  • 19 crates.io advisories · 1 critical · 5 high · 13 moderate
  • 18 RubyGems advisories · 1 critical · 13 high · 2 moderate · 2 low
  • 18 NuGet advisories · 2 critical · 6 high · 9 moderate · 1 low
  • 2 Multi-ecosystem advisories · 1 critical · 1 high
  • 1 GitHub Actions advisory · 1 high
  • 1 Hex 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.