arazzo: 1.0.1
info:
  title: Vercel AI Gateway Discover Model Then Complete
  summary: List available models, confirm the requested one is present, then run a chat completion.
  description: >-
    A guarded chat-completion flow for the Vercel AI Gateway. The workflow first
    lists the models reachable through the gateway and confirms the requested
    model id appears in the catalog, then issues a chat completion against it. This
    prevents spending a request against a model the account cannot reach. Each step
    inlines its request and documents the success status.
  version: 1.0.0
sourceDescriptions:
- name: aiGatewayApi
  url: ../openapi/vercel-ai-gateway-openapi.yml
  type: openapi
workflows:
- workflowId: discoverModelThenComplete
  summary: Verify a model is in the gateway catalog, then create a chat completion.
  description: >-
    Lists gateway models, branches on whether the requested model is available,
    and on success creates a chat completion returning the assistant message and
    token usage.
  inputs:
    type: object
    required:
    - apiKey
    - model
    - prompt
    properties:
      apiKey:
        type: string
        description: Vercel AI Gateway API key (AI_GATEWAY_API_KEY) used as a Bearer token.
      model:
        type: string
        description: "Model id in 'provider/model-name' form (e.g. anthropic/claude-opus-4.6)."
      prompt:
        type: string
        description: The user message to send to the model.
      maxTokens:
        type: integer
        description: Optional maximum number of tokens to generate.
  steps:
  - stepId: listModels
    description: List all models reachable through the gateway to confirm availability.
    operationId: listModels
    parameters:
    - name: Authorization
      in: header
      value: Bearer $inputs.apiKey
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      models: $response.body#/data
    onSuccess:
    - name: modelAvailable
      type: goto
      stepId: complete
      criteria:
      - context: $response.body
        condition: $.data[?(@.id == '$inputs.model')]
        type: jsonpath
    - name: modelMissing
      type: end
      criteria:
      - context: $response.body
        condition: $.data[?(@.id == '$inputs.model')].length == 0
        type: jsonpath
  - stepId: complete
    description: Create a chat completion against the confirmed model.
    operationId: createChatCompletion
    parameters:
    - name: Authorization
      in: header
      value: Bearer $inputs.apiKey
    requestBody:
      contentType: application/json
      payload:
        model: $inputs.model
        messages:
        - role: user
          content: $inputs.prompt
        max_tokens: $inputs.maxTokens
        stream: false
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      completionId: $response.body#/id
      reply: $response.body#/choices/0/message/content
      finishReason: $response.body#/choices/0/finish_reason
      totalTokens: $response.body#/usage/total_tokens
  outputs:
    reply: $steps.complete.outputs.reply
    finishReason: $steps.complete.outputs.finishReason
    totalTokens: $steps.complete.outputs.totalTokens