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, Friday 17 April 2026 Retrospective Covers 16 April 2026 00:00 to 17 April 2026 00:00 UTC Web: https://keryn.dev/headlines/daily/2026-04-17/ TOP DEVELOPMENTS [1] Cisco says critical Webex Services flaw requires customer action BleepingComputer · Analysis · https://bleepingcomputer.com/news/security/cisco-says-critical-webex-services-flaw-requires-customer-action [2] Don't Make Models Guess Security and Safety: Symbolic Guardrails for Domain-Specific AI Agents arXiv · Research · https://arxiv.org/abs/2604.15579 [3] Towards Reliable Testing of Machine Unlearning arXiv · Research · https://arxiv.org/abs/2604.16536 SAST [4] NFTDELTA: Detecting Permission Control Vulnerabilities in NFT Contracts through Multi-View Learning arXiv · Research · https://arxiv.org/abs/2604.15118 SCA [5] The case for dependency cooldowns in a post-axios world Datadog · Analysis · https://securitylabs.datadoghq.com/articles/dependency-cooldowns AGENTIC SDLC [6] The Semi-Executable Stack: Agentic Software Engineering and the Expanding Scope of SE arXiv · Research · https://arxiv.org/abs/2604.15468 REMEDIATION [7] Securing AI Applications From Inception to Deployment Wiz · Analysis · https://wiz.io/blog/securing-ai-application-from-inception-to-deployment [8] Microsoft: April Windows Server 2025 update may fail to install BleepingComputer · Analysis · https://bleepingcomputer.com/news/microsoft/microsoft-april-windows-server-2025-update-may-fail-to-install MODELS [9] Route to Rome Attack: Directing LLM Routers to Expensive Models via Adversarial Suffix Optimization arXiv · Research · https://arxiv.org/abs/2604.15022 [10] Benchmarks for Trajectory Safety Evaluation and Diagnosis in OpenClaw and Codex: ATBench-Claw and ATBench-Codex arXiv · Research · https://arxiv.org/abs/2604.14858 [11] Hijacking Large Audio-Language Models via Context-Agnostic and Imperceptible Auditory Prompt Injection arXiv · Research · https://arxiv.org/abs/2604.14604 AI SECURITY [12] HarmfulSkillBench: How Do Harmful Skills Weaponize Your Agents? arXiv · Research · https://arxiv.org/abs/2604.15415 [13] New ATHR vishing platform uses AI voice agents for automated attacks BleepingComputer · Analysis · https://bleepingcomputer.com/news/security/new-athr-vishing-platform-uses-ai-voice-agents-for-automated-attacks [14] Anumati: Proof of Adherence as a Formal Consent Model for Autonomous Agent Protocols arXiv · Research · https://arxiv.org/abs/2604.16524 ADVISORIES IN THIS WINDOW 63 npm advisories · 14 critical · 25 high · 22 moderate · 2 low 25 PyPI advisories · 2 critical · 5 high · 16 moderate · 2 low 12 Packagist advisories · 3 critical · 5 high · 4 moderate 10 Go advisories · 1 critical · 5 high · 4 moderate 4 Maven advisories · 2 high · 2 moderate 3 crates.io advisories · 3 low 3 NuGet advisories · 2 critical · 1 high 1 RubyGems advisory · 1 moderate 1 Multi-ecosystem advisory · 1 moderate All advisories: https://keryn.dev/headlines/advisories/ Titles, links and short excerpts only; every item links to its source. Preview: the unsubscribe link and Keryn's postal address are added to each email when it is sent.