AI Engineer

1,000 Agent Tasks in a Sandbox: What Breaks When LLMs Run Code — Kevin Orellana, AWS

Oct 12, 2026 · 0:15:19

Listen to Episode

✨ AI Summary

Kevin Orellana tested an AWS coding agent in a sandbox.
1,000 tasks were executed across 10 domains to find failures.
Results were logged to SQLite, detailing sandbox vulnerabilities.

Guests on This Episode

KO

Kevin Orellana

1 podcast appearance

More from AI Engineer

Build a Research Agent with Claude Managed Agents (Part 2 of 4) — Anthropic

Oct 12, 2026 · 0:35:36

Decoupling the Brain from the Hands: Claude Managed Agents (Part 1 of 4) — Anthropic

Oct 12, 2026 · 0:17:40

Inside the AI Economy: What Stripe's Data Reveals — Nilofer Rajpurkar, Stripe

Oct 12, 2026 · 0:20:50

Building AIE: Open Q&A and Feedback Session with swyx

Oct 12, 2026 · 0:58:42

View all episodes →