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, Saturday 17 October 2026 · Retrospective · Covers 16 October 2026 00:00 to 17 October 2026 00:00 UTC

Read on the web

Top developments

Case Study: GitHub Expands Malware Advisories Across Eight Ecosystems with OpenSSF

OpenSSF · Research · SCA

What We Can Learn About GitHub Actions Security from the Trivy Breach

Endor Labs · Incident · SCA

GitHub Action tj-actions/changed-files supply chain attack: what you need to know

Endor Labs · Incident · SCA

SAST

Does AI-Assisted Development Homogenize Software? An Empirical Study of Code Similarity, Quality, and Evolution

arXiv · Research · Agentic SDLC, SAST

VulnSAGE: Adjudicating SAST Reports by Connecting LLM Reasoning to Program Structure

arXiv · Research · SAST

SCA

An Interactive GraphRAG Workspace for Reifying, Auditing, and Steering Conversational AI Memories

arXiv · Research · AI Security, SCA

Container Scanning + SCA = Better Together

Endor Labs · Analysis · SCA

DAST

Efficient Program Reduction for Library Fuzzing Harnesses

arXiv · Research · DAST

Top Escape.tech alternatives for offensive security in 2026

Aikido Security · Analysis · DAST

Agentic SDLC

Who Verifies the Builder? Toward Independent Assurance for Agent-Produced Software

arXiv · Research · Agentic SDLC

Trojan Horse: Automated Repository Poisoning against Coding Agents

arXiv · Research · Agentic SDLC

From Approval to Abuse: Privilege Escalation Attacks on Harness-Protected Coding Agents

arXiv · Research · Agentic SDLC, AI Security

Remediation

MLOps-Bench: Benchmarking AI Agents on Production ML Engineering in Real Repositories

arXiv · Research · Remediation

Trust

GatewayBench: Unified Capability and Security Evaluation Platform for MCP Gateways

arXiv · Research · AI Security, Trust

SAFE-Mem: Secure Agentic Memory with Fine-Grained Authority and Governance-Aware Retrieval

arXiv · Research · Trust

Identify, Classify, Act: Per-Instruction Controlled Compliance for LLMs over External Data

arXiv · Research · AI Security, Models, Trust

Models

Under the Skin: Backdoor Infection in Multimodal Large Language Models

arXiv · Research · Models

Red-Teaming Multi-Agent Attacks for Loss-of-Control Risks

arXiv · Research · Models

Monitor-Validity Failures: LLM Safety Evidence Paths at Evolving Framework Boundaries

arXiv · Research · Models

AI Security

RATION: Towards Quantitative Risk-Based Access Control for AI Agents

arXiv · Research · AI Security

SPIDES: Information-Flow Control for Browser Agents

arXiv · Research · AI Security

The Window Is in the Reasoning: When LLM Agents Decide to Follow Prompt Injections

arXiv · Research · AI Security

Advisories in this window

  • 25 npm advisories · 24 critical · 1 low
  • 5 Go advisories · 3 high · 1 moderate · 1 low
  • 5 PyPI advisories · 4 critical · 1 low
  • 2 crates.io advisories · 1 high · 1 moderate
  • 1 Maven advisory · 1 high
  • 1 Hex advisory · 1 low

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.