Designing trust into agentic systems.

Six interaction patterns for AI agents. Written and built by Eric Sparks, a product designer at Microsoft.

Demo

What's the deadline for the Acme contract renewal?

The Acme contract renewal deadline is March 15, 2026. The original contract was signed on March 15, 2023 with a 3-year term.

Ask a follow-up

Library

Six patterns

Chapter one is live. The remaining chapters are currently in writing.

Confidence Indicators

Live

Show how certain the agent is before the user has to ask.

Open pattern

Thought Transparency

In writing

Let users see into the agent's reasoning without overwhelming them.

Override Controls

In writing

Make it effortless to correct or stop the agent mid-task.

Error Recovery

In writing

Turn agent failures into user trust moments.

Progressive Autonomy

In writing

Let users grant the agent more freedom as trust builds.

Intent Confirmation

In writing

Make the user's mapped intent explicit before execution.

Why this exists

The interaction patterns are not documented.

The engineering patterns for AI agents are well documented. The interaction patterns are not. How trust, control, and recovery should feel from the user's side of the screen is still spread across blog posts, conference talks, and product launches.

Agent Inspector names those patterns, builds them, and argues for them.

Recognized by

CSS Design Awards 2026

Best UI, Best UX & Best Innovation