Coverage topic

AI Agents

Autonomy, tools, orchestration, permissions and the product systems required to make agents useful.

October 26, 2026 · New analysis

Test the workflow behind the format.

Open Agent Standards Need a Portability Test

AAIF’s four October 22 additions cover procedures, model interfaces, training and execution. Product teams still need evidence that a migrated workflow behaves correctly.

↗

October 23, 2026 · New analysis

Fewer handoffs. Better questions.

All stories ↗

EngineeringOct 23

Your AI Agent May Need Fewer Agents

Anthropic’s October 22 guidance challenges unnecessary agent handoffs. A practical way to compare shared context, specialist workers and the outcomes users trust.

↗

October 22, 2026 · New analysis

When the agent comes to you.

All stories ↗

AI AgentsOct 22

Claude Duet Is Here. Your Attention Is the Interface.

The next AI interface decides when to interrupt you. Claude Duet makes that design problem concrete—and puts evidence, approval and attention on the product roadmap.

↗

October 2026 · New analysis

New in AI Agents.

All stories ↗

AI AgentsOct 12

Prompt injection is an interface problem, not only a model bug

No model reliably separates instructions from data. Agent products have to make that boundary visible, limit risky capability combinations and show where a request came from.

↗

Published Sep 12, 2026

From the desk

All stories ↗

AI AgentsSep 12

The Agents API changes the build-versus-orchestrate decision.

OpenAI is moving agent loops, context management and long-running execution behind an API. That shifts where product teams should spend engineering effort.

↗

AI AgentsSep 12

Agent harness portability is becoming infrastructure, not a nice-to-have.

Vercel’s AI SDK can now run GitHub Copilot through the same HarnessAgent interface as other coding agents. The abstraction is moving above the model.

↗