DRAFT — placeholder data. Replace with a real agent-tested trace before publishing.

Agent-friendly

Sample API (green)

We sent a deliberately malformed request. It read the error, fixed the payload, and retried — no human touched it.

The five tasks

Getting started from the docs aloneNailed it

Fixing its own mistake after an errorNailed it

Here’s the receipt — what actually happened, not our summary of it.

> POST /v1/things  { "naem": "demo" }
< 400  field "naem" is not recognized; did you mean "name"?
> POST /v1/things  { "name": "demo" }
< 201  created id=thing_8f2a

Following a multi-step flowNailed it

Handling an unclear edge caseHalf-nailed it

Not double-charging on a retryNailed it

See everything the AI did (3 steps)

12:00:01  Claude   GET /docs  → 200
12:00:04  Claude   POST /v1/things {naem:"demo"}  → 400
12:00:06  Claude   POST /v1/things {name:"demo"}  → 201

Tested 2026-06-19 with Claude, GPT, Gemini agents · request a re-test ↗