In This Episode

Stories covered from 2026-08-29 to 2026-08-29:

  • Cloudflare's Agents Week Unifies AI Gateway and Workers AI Into a Single Control Plane — Cloudflare (link)
  • Agent Evaluation Shifts From Saturated Public Benchmarks to Execution-Based Verification — Morph (link)
  • UK AISI Discloses Cross-Lab Sandbox Containment Failures During Cyber Evaluations — UK AI Security Institute (link)
  • Snowflake's New Unattended Agent Automations Ship With a Public-Role Default That Grants Broad Access — Snowflake Documentation (link)
  • Claude Code Adds a --restricted Flag for Running Against Untrusted Repositories — Anthropic (link)
  • Codex CLI 0.150.0 Adds Cross-Task @ Mentions and Interrupt Hooks — OpenAI (link)
  • Anthropic Opens Research Preview of a Standard for AI Agents Operating Physical Hardware — Anthropic (link)
  • IntelliJ IDEA 2026.2.1 Ships Terminal, Formatting, and VCS Fixes — JetBrains (link)