Sandbox Breakouts: Rapid Unscheduled Agentic Assembly by Andrew Martin
Andrew Martin, CEO @controlplane584 explores what really happens inside AI agent sandboxes, where coding assistants can write and execute code, access files, and interact with networks on your behalf. In this talk, Andrew walks through how his team built probes that run inside these environments to analyze their security posture from an attacker’s perspective. By comparing sandboxed behavior to unsandboxed baselines, they uncovered gaps, leaks, and assumptions that do not always hold up in practice. This session focuses on real findings and practical approaches to isolating agent execution using cloud native tools. If you are working with AI agents, platform security, or Kubernetes-based systems, this talk offers a grounded look at the risks and what you can actually do about them. Recorded live at KCD Texas 2026 on May 15th, in Austin, Texas. Kubernetes Community Days Texas is a community-driven event focused on sharing knowledge and bringing together people working across the cloud native ecosystem. Thank you to our sponsors for making this event possible: https://kcdtexas.org Follow KCD Texas: @KCDTexas

Attacking AI - Jason Haddix - NDC Security 2026

Enhance Your Agent Accuracy with Agent Skills by Lin Sun

Security Assessment A→Z: The Complete Methodology (Full 54-Min Training)

Andrej Karpathy: From Vibe Coding to Agentic Engineering w/ Stephanie Zhan

Stop Wasting GPUs: How We Built a Golden Path for GPU Sharing on Kubernetes by Engin Diri

Livin' In the Future: Your Platform's Next Interface Is an AI Agent by Whitney Lee

Linus Torvalds: AI Is Changing Linux Fast

Zig 2026: No-AI Policy, $670K Foundation, Left GitHub & Why Zig Isn’t 1.0 - Andrew Kelley Explains

How did that happen? And is it a security Problem by Duffie Cooley

The Most Mysterious File On The Internet

Trump’s Childish Behavior with World Leaders, Republicans Bash His Iran Deal & Guillermo’s Huge News

OWASP's Top 10 Ways to Attack LLMs: AI Vulnerabilities Exposed

DEF CON 33 - Exploiting Shadow Data from AI Models and Embeddings - Patrick Walsh

Mastering the Dark Arts of OTel Debugging by Julia Furst Morgado

5 CLI Tools That Actually Changed How I Work in 2026

Self-Service, Multi-Tenant Infrastructure With Kured and Flatcar Linux by Tyler Auerbeck

The most rational take on AI you’ll hear this year

Time to switch?

Why AI Can Never Escape Turing's 1936 Proof

