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

Daily, Wednesday 17 June 2026 · Retrospective · Covers 16 June 2026 00:00 to 17 June 2026 00:00 UTC

Read on the web

Top developments

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

Ars Technica: Security · Research · Agentic SDLC, Remediation

A Forgotten Contributor Account Compromised the Entire Mastra npm Package Scope

Snyk · Incident · SCA

CISA warns of another cPanel plugin flaw exploited in attacks

BleepingComputer · Incident

AI-BOM

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

arXiv · Research · AI-BOM, Trust

Remediation

The Fable 5 Export Controls Harm US Cyber Defense

Simon Willison · Analysis · Models, Remediation

What are git worktrees, and why should I use them?

GitHub · Analysis · Remediation

Attackers Exploit Three Fortinet FortiSandbox Flaws, One Patched Last Week

The Hacker News · Analysis · Remediation

Trust

Agentra: A Supervisable Multi-Agent Framework for Enterprise Intrusion Response

arXiv · Research · Trust

AI Transparency: Governance Compliance or Stakeholder Requirements?

arXiv · Research · Trust

Models

A Red-Team Study of Anthropic Fable 5 & Opus 4.8 Models

arXiv · Research · Models

Differential Privacy of Gaussian Process Posterior Sampling

arXiv · Research · Models

AnchorKV: Safety-Aware KV Cache Compression via Soft Penalty with a Refusal Anchor

arXiv · Research · Models

AI Security

SafeClawBench: Separating Semantic, Audit-Evidence, and Sandbox Harm in Tool-Using LLM Agents

arXiv · Research · AI Security

MIRAGE: Stealthy Visual Prompt Injection for Vulnerability Detection in Web Agents

arXiv · Research · AI Security

PARSE: Provenance-Aware Retrieval Sanitization for Professional Domain LLM Agents

arXiv · Research · AI Security

Advisories in this window

  • 74 npm advisories · 40 critical · 12 high · 20 moderate · 2 low
  • 25 PyPI advisories · 4 critical · 9 high · 11 moderate · 1 low
  • 16 Go advisories · 10 high · 6 moderate
  • 8 crates.io advisories · 3 high · 5 moderate
  • 1 Packagist advisory · 1 moderate
  • 1 NuGet 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.