# Generated by API Evangelist (build-phrasing.py). Our phrasing, not observed demand.
overlay: 1.0.0
info:
  title: API Evangelist conversational phrasing for Microsoft Azure Image Prediction API
  version: 1.0.0
extends: openapi/microsoft-azure-imagepredictionapi-api-openapi.yml
actions:
- target: $.info
  update:
    x-apievangelist-phrasing:
      method: generated
      generated: '2026-09-26'
      generator: build-phrasing.py
      label: Generated by API Evangelist
      operations: 12
- target: $.paths['/{projectId}/classify/iterations/{publishedName}/url'].post
  update:
    x-apievangelist-phrasing:
      intent: Classify an image from a URL and save the result
      effect: write
      questions:
      - How do I classify an image hosted at a URL with my published Custom Vision model and keep the prediction?
      - Can I tag a web image with my classifier and have the prediction stored in the project?
      instructions:
      - text: Classify the image at {url} with iteration {publishedName} of project {projectId} and save the result.
        slots:
          url: requestBody.url
          publishedName: path.publishedName
          projectId: path.projectId
      - text: Run classifier {publishedName} on image URL {url} for project {projectId}, storing the prediction under application {application}.
        slots:
          publishedName: path.publishedName
          url: requestBody.url
          projectId: path.projectId
          application: query.application
      method: generated
      generated: '2026-09-26'
- target: $.paths['/{projectId}/classify/iterations/{publishedName}/image'].post
  update:
    x-apievangelist-phrasing:
      intent: Classify an uploaded image and save the result
      effect: write
      questions:
      - How do I send image bytes to my Custom Vision classifier and keep the prediction?
      - Can I upload a local photo for classification and have it saved to the project?
      instructions:
      - text: Classify uploaded image {imageData} with iteration {publishedName} of project {projectId} and save the result.
        slots:
          imageData: requestBody.imageData
          publishedName: path.publishedName
          projectId: path.projectId
      - text: Upload this file {imageData} to classifier {publishedName} in project {projectId} and keep the prediction.
        slots:
          imageData: requestBody.imageData
          publishedName: path.publishedName
          projectId: path.projectId
      method: generated
      generated: '2026-09-26'
- target: $.paths['/{projectId}/classify/iterations/{publishedName}/url/nostore'].post
  update:
    x-apievangelist-phrasing:
      intent: Classify an image URL without storing it
      effect: read
      questions:
      - Can I classify an image at a URL without the image or prediction being saved?
      - Which call classifies a web image privately, leaving nothing stored in the project?
      instructions:
      - text: Classify the image at {url} with iteration {publishedName} of project {projectId} without saving it.
        slots:
          url: requestBody.url
          publishedName: path.publishedName
          projectId: path.projectId
      - text: Run a no-store classification of image URL {url} on project {projectId} iteration {publishedName}.
        slots:
          url: requestBody.url
          projectId: path.projectId
          publishedName: path.publishedName
      method: generated
      generated: '2026-09-26'
- target: $.paths['/{projectId}/classify/iterations/{publishedName}/image/nostore'].post
  update:
    x-apievangelist-phrasing:
      intent: Classify an uploaded image without storing it
      effect: read
      questions:
      - Can I classify uploaded image bytes without Custom Vision keeping a copy?
      - What's the way to classify a local file without the result being saved?
      instructions:
      - text: Classify uploaded image {imageData} with iteration {publishedName} of project {projectId} without saving it.
        slots:
          imageData: requestBody.imageData
          publishedName: path.publishedName
          projectId: path.projectId
      - text: Run a no-store classification of file {imageData} on project {projectId} iteration {publishedName}.
        slots:
          imageData: requestBody.imageData
          projectId: path.projectId
          publishedName: path.publishedName
      method: generated
      generated: '2026-09-26'
- target: $.paths['/{projectId}/detect/iterations/{publishedName}/url'].post
  update:
    x-apievangelist-phrasing:
      intent: Detect objects in an image URL and save the result
      effect: write
      questions:
      - How do I find and locate objects in an image at a URL using my object detection model?
      - Can I get bounding boxes for a web image and keep the detection in the project?
      instructions:
      - text: Detect objects in the image at {url} with iteration {publishedName} of project {projectId} and save the result.
        slots:
          url: requestBody.url
          publishedName: path.publishedName
          projectId: path.projectId
      - text: Run object detector {publishedName} on image URL {url} for project {projectId}, storing it under application {application}.
        slots:
          publishedName: path.publishedName
          url: requestBody.url
          projectId: path.projectId
          application: query.application
      method: generated
      generated: '2026-09-26'
- target: $.paths['/{projectId}/detect/iterations/{publishedName}/image'].post
  update:
    x-apievangelist-phrasing:
      intent: Detect objects in an uploaded image and save the result
      effect: write
      questions:
      - How do I upload a photo and get the objects my detection model finds in it, saved to the project?
      - Can I run object detection on local image bytes and keep the result?
      instructions:
      - text: Find objects in the photo bytes {imageData} I upload, using detector {publishedName} of project {projectId}, and store the detections.
        slots:
          imageData: requestBody.imageData
          publishedName: path.publishedName
          projectId: path.projectId
      - text: Upload file {imageData} to object detector {publishedName} in project {projectId} and keep the detections.
        slots:
          imageData: requestBody.imageData
          publishedName: path.publishedName
          projectId: path.projectId
      method: generated
      generated: '2026-09-26'
- target: $.paths['/{projectId}/detect/iterations/{publishedName}/url/nostore'].post
  update:
    x-apievangelist-phrasing:
      intent: Detect objects in an image URL without storing it
      effect: read
      questions:
      - Can I detect objects in a web image without the image or detections being saved?
      - Which call finds objects at an image URL and leaves nothing in my project?
      instructions:
      - text: Detect objects in the image at {url} with iteration {publishedName} of project {projectId} without saving it.
        slots:
          url: requestBody.url
          publishedName: path.publishedName
          projectId: path.projectId
      - text: Run a no-store object detection on image URL {url} for project {projectId} iteration {publishedName}.
        slots:
          url: requestBody.url
          projectId: path.projectId
          publishedName: path.publishedName
      method: generated
      generated: '2026-09-26'
- target: $.paths['/{projectId}/detect/iterations/{publishedName}/image/nostore'].post
  update:
    x-apievangelist-phrasing:
      intent: Detect objects in an uploaded image without storing it
      effect: read
      questions:
      - Can I run object detection on an uploaded image without Custom Vision keeping it?
      - What's the way to detect objects in a local file privately?
      instructions:
      - text: Detect objects in uploaded image {imageData} with iteration {publishedName} of project {projectId} without saving it.
        slots:
          imageData: requestBody.imageData
          publishedName: path.publishedName
          projectId: path.projectId
      - text: Run a no-store object detection on file {imageData} for project {projectId} iteration {publishedName}.
        slots:
          imageData: requestBody.imageData
          projectId: path.projectId
          publishedName: path.publishedName
      method: generated
      generated: '2026-09-26'
- target: $.paths['/{projectId}/url'].post
  update:
    x-apievangelist-phrasing:
      intent: Predict an image URL with a prediction key (older API)
      effect: write
      questions:
      - How do I get a prediction for an image URL using a prediction key and an iteration id?
      - Can the older predict endpoint score a web image against a specific iteration and save it?
      instructions:
      - text: Predict the image at {Url} for project {projectId} using prediction key {Prediction-Key} and save the result.
        slots:
          Url: requestBody.Url
          projectId: path.projectId
          Prediction-Key: header.Prediction-Key
      - text: Score image URL {Url} against iteration {iterationId} of project {projectId} with the older predict endpoint.
        slots:
          Url: requestBody.Url
          iterationId: query.iterationId
          projectId: path.projectId
      method: generated
      generated: '2026-09-26'
- target: $.paths['/{projectId}/image'].post
  update:
    x-apievangelist-phrasing:
      intent: Predict an uploaded image with a prediction key (older API)
      effect: write
      questions:
      - How do I get a prediction for uploaded image bytes by iteration id using a prediction key?
      - Can the older predict endpoint score a local image and store the result?
      instructions:
      - text: Predict uploaded image {imageData} for project {projectId} using prediction key {Prediction-Key} and save the result.
        slots:
          imageData: requestBody.imageData
          projectId: path.projectId
          Prediction-Key: header.Prediction-Key
      - text: Score file {imageData} against iteration {iterationId} of project {projectId} with the older predict endpoint.
        slots:
          imageData: requestBody.imageData
          iterationId: query.iterationId
          projectId: path.projectId
      method: generated
      generated: '2026-09-26'
- target: $.paths['/{projectId}/url/nostore'].post
  update:
    x-apievangelist-phrasing:
      intent: Predict an image URL without storing it (older API)
      effect: read
      questions:
      - Can the older predict endpoint score an image URL by iteration id without saving anything?
      - Which prediction-key call evaluates a web image and keeps no copy?
      instructions:
      - text: Predict the image at {Url} for project {projectId} using prediction key {Prediction-Key} without saving it.
        slots:
          Url: requestBody.Url
          projectId: path.projectId
          Prediction-Key: header.Prediction-Key
      - text: Run a no-store prediction on image URL {Url} against iteration {iterationId} of project {projectId}.
        slots:
          Url: requestBody.Url
          iterationId: query.iterationId
          projectId: path.projectId
      method: generated
      generated: '2026-09-26'
- target: $.paths['/{projectId}/image/nostore'].post
  update:
    x-apievangelist-phrasing:
      intent: Predict an uploaded image without storing it (older API)
      effect: read
      questions:
      - Can the older predict endpoint score uploaded image bytes without keeping them?
      - Which prediction-key call evaluates a local file privately by iteration id?
      instructions:
      - text: Predict uploaded image {imageData} for project {projectId} using prediction key {Prediction-Key} without saving it.
        slots:
          imageData: requestBody.imageData
          projectId: path.projectId
          Prediction-Key: header.Prediction-Key
      - text: Run a no-store prediction on file {imageData} against iteration {iterationId} of project {projectId}.
        slots:
          imageData: requestBody.imageData
          iterationId: query.iterationId
          projectId: path.projectId
      method: generated
      generated: '2026-09-26'