Zywrap
Interactive demo

Don't guess the output. See the code.

Explore real wrapper-code examples, inspect the request payload, and run a simulated response before moving the same API shape into the playground.

Open playground API reference

Workflow lifecycle

Input fields or files

Wrapper code selects behavior

Run executes on Zywrap

Structured output returns

Artifact and usage record saved

Marketing Copywriting

Google RSA Generator

Generate compliant Google Responsive Search Ad headlines and descriptions for a target keyword and audience.

mc_google_responsive_search_ads_rsa_base
POST /v2/runs

Request payload

{
  "model": "openai-gpt-4o-mini",
  "wrapperCodes": [
    "mc_google_responsive_search_ads_rsa_base"
  ],
  "input": {
    "text": "Generate compliant Google Responsive Search Ad headlines and descriptions for a target keyword and audience.",
    "variables": {
      "productName": "Zywrap",
      "audience": "SaaS developers",
      "primaryKeyword": "AI workflow API",
      "offer": "Build AI features without prompt engineering"
    }
  },
  "output": {
    "format": "json",
    "artifact": false
  }
}

Response

Click “Run request” to preview the structured output.

Demo examples use the same primitives your API uses.

The request shape matches Zywrap workflow execution: wrapperCodes, model selection, input variables, files, artifacts, and usage records.

Prompt logic remains server-side.

Outputs are traceable through runs and ledger records.