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 25 March 2026 Retrospective Covers 24 March 2026 00:00 to 25 March 2026 00:00 UTC Web: https://keryn.dev/headlines/daily/2026-03-25/ TOP DEVELOPMENTS [1] LiteLLM and Telnyx compromised on PyPI: Tracing the TeamPCP supply chain campaign Datadog · Incident · https://securitylabs.datadoghq.com/articles/litellm-compromised-pypi-teampcp-supply-chain-campaign [2] Popular LiteLLM PyPI package backdoored to steal credentials, auth tokens BleepingComputer · Incident · https://bleepingcomputer.com/news/security/popular-litellm-pypi-package-compromised-in-teampcp-supply-chain-attack [3] Ghost Campaign Uses 7 npm Packages to Steal Crypto Wallets and Credentials The Hacker News · Incident · https://thehackernews.com/2026/03/ghost-campaign-uses-7-npm-packages-to.html [4] TeamPCP Hacks Checkmarx GitHub Actions Using Stolen CI Credentials The Hacker News · Incident · https://thehackernews.com/2026/03/teampcp-hacks-checkmarx-github-actions.html [5] Citrix Urges Patching Critical NetScaler Flaw Allowing Unauthenticated Data Leaks The Hacker News · Incident · https://thehackernews.com/2026/03/citrix-urges-patching-critical.html SAST [6] TreeTeaming: Autonomous Red-Teaming of Vision-Language Models via Hierarchical Strategy Exploration arXiv · Research · https://arxiv.org/abs/2603.22882 [7] Agent Audit: A Security Analysis System for LLM Agent Applications arXiv · Research · https://arxiv.org/abs/2603.22853 [8] Does Teaming-Up LLMs Improve Secure Code Generation? A Comprehensive Evaluation with Multi-LLMSecCodeEval arXiv · Research · https://arxiv.org/abs/2603.22717 SCA [9] You were one pip install away from the litellm breach. Chainguard customers weren’t. Chainguard · Incident · https://chainguard.dev/unchained/you-were-one-pip-install-away-from-a-breach-chainguard-customers-werent [10] Package Managers Need to Cool Down Simon Willison · Analysis · https://simonwillison.net/2026/Mar/24/package-managers-need-to-cool-down DAST [11] Towards Leveraging LLMs to Generate Abstract Penetration Test Cases from Software Architecture arXiv · Research · https://arxiv.org/abs/2603.23698 [12] Checkmarx DAST for the AI Coding Era: Runtime Security at Machine Speed Checkmarx · Analysis · https://checkmarx.com/blog/checkmarx-dast-for-the-ai-coding-era AGENTIC SDLC [13] Auto mode for Claude Code Simon Willison · Analysis · https://simonwillison.net/2026/Mar/24/auto-mode-for-claude-code [14] Code Review Agent Benchmark arXiv · Research · https://arxiv.org/abs/2603.23448 [15] Supercharge Your SOC Elastic · Analysis · https://elastic.co/security-labs/blog/supercharge-your-soc REMEDIATION [16] Building AI-powered GitHub issue triage with the Copilot SDK GitHub · Analysis · https://github.blog/ai-and-ml/github-copilot/building-ai-powered-github-issue-triage-with-the-copilot-sdk TRUST [17] Governing AI agent behavior: Aligning user, developer, role, and organizational intent Microsoft · Analysis · https://techcommunity.microsoft.com/blog/microsoft-security-blog/governing-ai-agent-behavior-aligning-user-developer-role-and-organizational-inte/4503551 [18] Privacy-Preserving EHR Data Transformation via Geometric Operators: A Human-AI Co-Design Technical Report arXiv · Research · https://arxiv.org/abs/2603.22954 [19] From the AI Act to a European AI Agency: Completing the Union's Regulatory Architecture arXiv · Research · https://arxiv.org/abs/2603.22912 MODELS [20] Not All Tokens Are Created Equal: Query-Efficient Jailbreak Fuzzing for LLMs arXiv · Research · https://arxiv.org/abs/2603.23269 [21] SafeSeek: Universal Attribution of Safety Circuits in Language Models arXiv · Research · https://arxiv.org/abs/2603.23268 [22] RTS-ABAC: Real-Time Server-Aided Attribute-Based Authorization & Access Control for Substation Automation Systems arXiv · Research · https://arxiv.org/abs/2603.23012 AI SECURITY [23] The Cognitive Firewall:Securing Browser Based AI Agents Against Indirect Prompt Injection Via Hybrid Edge Cloud Defense arXiv · Research · https://arxiv.org/abs/2603.23791 [24] Mind Your HEARTBEAT! Claw Background Execution Inherently Enables Silent Memory Pollution arXiv · Research · https://arxiv.org/abs/2603.23064 [25] AgentRAE: Remote Action Execution through Notification-based Visual Backdoors against Screenshots-based Mobile GUI Agents arXiv · Research · https://arxiv.org/abs/2603.23007 ADVISORIES IN THIS WINDOW 26 npm advisories · 19 critical · 4 high · 2 moderate · 1 low 21 Go advisories · 5 high · 15 moderate · 1 low 9 Packagist advisories · 2 high · 5 moderate · 2 low 8 NuGet advisories · 1 critical · 5 high · 2 moderate 4 PyPI advisories · 3 high · 1 moderate 4 Maven advisories · 1 high · 2 moderate · 1 low 1 RubyGems advisory · 1 moderate 1 Multi-ecosystem advisory · 1 critical 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.