Stack layer / Threat pattern
Pick your agent sandbox primitive by threat model: Firecracker microVM vs gVisor vs V8 isolate
Northflank
Stack layer / Threat pattern
Microsoft Agent Framework python-1.11.0 adds Progressive MCP Disclosure and GA Skills API
GitHub (Microsoft Agent Framework)
Policy dependency / Stack layer
Pattern: A Provider-Abstraction / BYOK Layer Is Hardening Beneath Agent Harnesses
GitHub
Policy dependency / Stack layer
Make context compaction a trained policy, not an inference-time heuristic — CompactionRL lifts SWE-bench up to +7 points
arXiv 2607.05378
Stack layer / Update thread
EvoAgentBench: A Benchmark That Isolates Whether Agents Can Turn Past Sessions Into Reusable Procedures
arXiv
Policy dependency / Stack layer
Vercel Ships the Agent-Ops Layer: 'Eve' Framework for Natural-Language Agent Skills and 'Vercel Sandbox' to Fence In Data Access
The AI Insider
Stack layer / Update thread
OpenAI's GPT-Live Adds Async Delegation: Voice Keeps Talking While GPT-5.5 Works in the Background
Simon Willison
Stack layer / Threat pattern
Unit 42: AI-agent npm packages become a prime supply-chain target in H1 2026
Unit 42 (Palo Alto Networks)