ChatGPT Describe an Image Input

arazzo: 1.0.1
info:
  title: ChatGPT Describe an Image Input
  summary: Send an image URL to the Responses API and retrieve a text description.
  description: >-
    The OpenAI image generation and audio endpoints are not present in these
    specifications, so this workflow adapts the multimodal theme to what the
    Responses API actually supports: image inputs. An input_image content part
    carrying an image URL is sent alongside a text instruction, the response is
    polled to completion, and the generated description text is returned. Every
    step spells out its request inline so the flow can be read and executed
    without opening the underlying OpenAPI description.
  version: 1.0.0
  x-realizes-capability-ids:
  - BC-610.60
  x-capability-derivation:
    method: 'deterministic join: sourceDescriptions -> per-tag OpenAPI -> tag/capability edge. No classification at this step.'
    min_confidence: 0.7
    sources:
    - capability_id: BC-610.60
      capability_name: Artificial Intelligence Management
      spec: chatgpt-responses-api-openapi.yml
      confidence: 0.7
    model: Turbo EA Capabilities by Vincent Verdet — Turbo EA, https://github.com/vincentmakes/turbo-ea-capabilities, CC BY 4.0
sourceDescriptions:
- name: responsesApi
  url: ../openapi/chatgpt-responses-api-openapi.yml
  type: openapi
workflows:
- workflowId: image-input-describe
  summary: Describe an image supplied by URL using a multimodal Responses API call.
  description: >-
    Creates a response from a multimodal input combining an instruction and an
    input_image content part, polls to completion, and returns the description
    text.
  inputs:
    type: object
    required:
    - apiKey
    - model
    - imageUrl
    - instruction
    properties:
      apiKey:
        type: string
        description: OpenAI API key used as the Bearer credential.
      model:
        type: string
        description: Vision-capable model ID (e.g. gpt-4o).
      imageUrl:
        type: string
        description: The URL of the image to describe.
      instruction:
        type: string
        description: The text instruction guiding the description.
  steps:
  - stepId: describeImage
    description: >-
      Create a stored response with a multimodal input that pairs the text
      instruction with an input_image content part.
    operationId: createResponse
    parameters:
    - name: Authorization
      in: header
      value: "Bearer $inputs.apiKey"
    requestBody:
      contentType: application/json
      payload:
        model: $inputs.model
        input:
        - role: user
          content:
          - type: input_text
            text: $inputs.instruction
          - type: input_image
            image_url: $inputs.imageUrl
            detail: auto
        store: true
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      responseId: $response.body#/id
      status: $response.body#/status
    onSuccess:
    - name: needsPolling
      type: goto
      stepId: pollDescription
      criteria:
      - context: $response.body
        condition: $.status == "in_progress"
        type: jsonpath
    - name: alreadyDone
      type: goto
      stepId: retrieveDescription
      criteria:
      - context: $response.body
        condition: $.status == "completed"
        type: jsonpath
  - stepId: pollDescription
    description: >-
      Poll the response until image understanding finishes and it leaves the
      in_progress status.
    operationId: getResponse
    parameters:
    - name: Authorization
      in: header
      value: "Bearer $inputs.apiKey"
    - name: response_id
      in: path
      value: $steps.describeImage.outputs.responseId
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      status: $response.body#/status
    onSuccess:
    - name: keepPolling
      type: goto
      stepId: pollDescription
      criteria:
      - context: $response.body
        condition: $.status == "in_progress"
        type: jsonpath
    - name: settled
      type: goto
      stepId: retrieveDescription
      criteria:
      - context: $response.body
        condition: $.status != "in_progress"
        type: jsonpath
  - stepId: retrieveDescription
    description: >-
      Retrieve the settled response and extract the description text and token
      usage, including the image URL in the returned items.
    operationId: getResponse
    parameters:
    - name: Authorization
      in: header
      value: "Bearer $inputs.apiKey"
    - name: response_id
      in: path
      value: $steps.describeImage.outputs.responseId
    - name: include
      in: query
      value:
      - message.input_image.image_url
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      finalStatus: $response.body#/status
      descriptionText: $response.body#/output/0/content/0/text
      totalTokens: $response.body#/usage/total_tokens
  outputs:
    responseId: $steps.describeImage.outputs.responseId
    descriptionText: $steps.retrieveDescription.outputs.descriptionText
    finalStatus: $steps.retrieveDescription.outputs.finalStatus
    totalTokens: $steps.retrieveDescription.outputs.totalTokens