Miatz playground

Schema Cage Match

Turn off JSON enforcement, watch a mock app crash.

Coming to the Build-LabAIE-204BYOK AIAI-Concept-Vizdev · data

The Demystify signature

Renders the actual downstream crash one malformed field causes -- 'validate at the boundary' stops being an abstract rule and becomes a broken UI the learner watches happen live.

How it works

What goes in, what comes out

What it does

The same request fires at a model with schema enforcement off in one lane and on in the other; the off lane's malformed or hallucinated JSON visibly breaks a mock downstream app with an animated error cascade, while the on lane holds the contract. A 'break it' button lets the learner adversarially prompt the on lane to test whether constrained decoding actually holds.

You bring

A schema pick/edit plus a prompt.

You get

Two parallel JSON outputs, a pass/fail against the schema, and a downstream-app crash animation on failure.

You control

Schema editor (order object, function args, extraction-schema presets); enforcement on/off; adversarial-prompt toggle.

Where it's used

Demo/lab companion inside AIE-204.

JSON schema enforcementfunction callingvalidation at the boundary

Optional · watch & try

Go deeper, elsewhere

Hand-picked public explainers and open tools that complement this one — always optional, never required, never graded.

Structured Outputs with Pydantic & OpenAI Function CallingShows schema-constrained output enforcement in practice Anthropic function calling for structured LLM outputsSame pattern demonstrated against Claude's tool-use API Benefits of Function Calling with Structured Outputs60-second why-it-matters framing · ~1 min OpenAI function calling guideOfficial reference for JSON-schema-constrained tool calling Claude tool use docsOfficial Anthropic reference for defining and validating tool schemas

Concepts click when you open the machinery.

Three labs are already live and free — the same hands-on style this playground brings to its module.

Play the free labs Take the DSAT