# Generated by API Evangelist (build-phrasing.py). Our phrasing, not observed demand.
overlay: 1.0.0
info:
  title: API Evangelist conversational phrasing for Edge Impulse Raw data API
  version: 1.0.0
extends: openapi/edge-impulse-raw-data-api-openapi.yml
actions:
- target: $.info
  update:
    x-apievangelist-phrasing:
      method: generated
      generated: '2026-10-01'
      generator: build-phrasing.py
      label: Generated by API Evangelist
      operations: 66
- target: $.paths['/api/{projectId}/raw-data'].get
  update:
    x-apievangelist-phrasing:
      intent: List data samples in a dataset category
      effect: read
      questions:
      - What samples are in the training set of my Edge Impulse project?
      - Can I page through test samples filtered by label or filename?
      - Which samples fall between a minimum and maximum length?
      instructions:
      - text: List the {category} samples in project {projectId}.
        slots:
          category: query.category
          projectId: path.projectId
      - text: List {category} samples labeled {labels} in project {projectId}, {limit} at a time.
        slots:
          category: query.category
          labels: query.labels
          projectId: path.projectId
          limit: query.limit
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/count'].get
  update:
    x-apievangelist-phrasing:
      intent: Count data samples in a dataset category
      effect: read
      questions:
      - How many training samples does my project have?
      - What's the number of test samples carrying a particular label?
      instructions:
      - text: Count the {category} samples in project {projectId}.
        slots:
          category: query.category
          projectId: path.projectId
      - text: Tell me how many {category} samples in project {projectId} are labeled {labels}.
        slots:
          category: query.category
          projectId: path.projectId
          labels: query.labels
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/label-object-detection-queue'].get
  update:
    x-apievangelist-phrasing:
      intent: Get images waiting for bounding-box labeling
      effect: read
      questions:
      - Which images still need bounding boxes drawn in my object detection project?
      - Where do I find the unlabeled items in the labeling queue?
      instructions:
      - text: List the unlabeled items in the object detection queue for project {projectId}.
        slots:
          projectId: path.projectId
      - text: Show the images awaiting bounding-box labels in project {projectId}.
        slots:
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/label-object-detection-queue/count'].get
  update:
    x-apievangelist-phrasing:
      intent: Count items in the object detection labeling queue
      effect: read
      questions:
      - How many images are left to label in my object detection queue?
      - Is my bounding-box labeling queue empty yet?
      instructions:
      - text: Count the unlabeled items in the object detection queue for project {projectId}.
        slots:
          projectId: path.projectId
      - text: Tell me how many images remain in the labeling queue of project {projectId}.
        slots:
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/rebalance'].post
  update:
    x-apievangelist-phrasing:
      intent: Rebalance train/test split (original endpoint)
      effect: write
      questions:
      - How do I reset my dataset to an 80/20 train-test split using the original rebalance endpoint?
      - Will the older rebalance call reassign every sample deterministically?
      instructions:
      - text: Rebalance project {projectId} to 80/20 train/test using the original v1 rebalance endpoint.
        slots:
          projectId: path.projectId
      - text: Run the legacy dataset rebalance on project {projectId}.
        slots:
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/clear-all-object-detection-labels'].post
  update:
    x-apievangelist-phrasing:
      intent: Clear all bounding-box labels in a project
      effect: destructive
      questions:
      - How do I wipe every bounding box in my dataset and start labeling over?
      - Can I send all images back to the labeling queue in one go?
      instructions:
      - text: Clear all object detection labels in project {projectId}.
        slots:
          projectId: path.projectId
      - text: Remove every bounding box from project {projectId} and requeue all images for labeling.
        slots:
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/delete-all'].post
  update:
    x-apievangelist-phrasing:
      intent: Delete all samples in a project
      effect: destructive
      questions:
      - How do I delete my entire dataset across training and testing?
      - Does removing all samples also invalidate my DSP and learn blocks?
      instructions:
      - text: Delete every sample in project {projectId} across all categories.
        slots:
          projectId: path.projectId
      - text: Wipe the whole dataset of project {projectId}.
        slots:
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/delete-all/{category}'].post
  update:
    x-apievangelist-phrasing:
      intent: Delete all samples in one category
      effect: destructive
      questions:
      - Can I delete just the test set while keeping training data?
      - How do I remove every sample in a single dataset category?
      instructions:
      - text: Delete all {category} samples in project {projectId}.
        slots:
          category: path.category
          projectId: path.projectId
      - text: Clear out the {category} category of project {projectId}.
        slots:
          category: path.category
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/metadata'].get
  update:
    x-apievangelist-phrasing:
      intent: Get metadata for all samples in a category
      effect: read
      questions:
      - What metadata is attached to every training sample in my project?
      - Can I export the metadata of all test samples at once?
      instructions:
      - text: Get metadata for all {category} samples in project {projectId}.
        slots:
          category: query.category
          projectId: path.projectId
      - text: Show sample metadata across the {category} set of project {projectId}.
        slots:
          category: query.category
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/imported-from'].get
  update:
    x-apievangelist-phrasing:
      intent: List samples with 'imported from' metadata
      effect: read
      questions:
      - Which items from my data source have already been imported into the project?
      - Can I see every sample tagged with an 'imported from' origin?
      instructions:
      - text: List samples with an imported-from key in project {projectId}.
        slots:
          projectId: path.projectId
      - text: Show {limit} already-imported data items in project {projectId}.
        slots:
          limit: query.limit
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/{sampleId}'].get
  update:
    x-apievangelist-phrasing:
      intent: Get a single data sample
      effect: read
      questions:
      - What are the sensor values and label of a specific sample?
      - Can I fetch one sample but cap the number of payload values returned?
      instructions:
      - text: Get sample {sampleId} from project {projectId}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
      - text: Fetch sample {sampleId} in project {projectId} limited to {limitPayloadValues} values.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
          limitPayloadValues: query.limitPayloadValues
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/{sampleId}'].delete
  update:
    x-apievangelist-phrasing:
      intent: Delete a single sample
      effect: destructive
      questions:
      - How do I remove one bad sample from my dataset?
      - Does deleting a sample also remove it from cold storage?
      instructions:
      - text: Delete sample {sampleId} from project {projectId}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
      - text: Remove the single sample {sampleId} in project {projectId}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/{sampleId}/raw'].get
  update:
    x-apievangelist-phrasing:
      intent: Download a sample in its original format
      effect: read
      questions:
      - Can I download a sample exactly as it was uploaded to the ingestion service?
      - How do I get the original file behind a sample?
      instructions:
      - text: Download the original uploaded file for sample {sampleId} in project {projectId}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
      - text: Get sample {sampleId} from project {projectId} in its raw uploaded format.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/{sampleId}/wav'].get
  update:
    x-apievangelist-phrasing:
      intent: Get an audio sample as a WAV file
      effect: read
      questions:
      - How do I listen to an audio sample as a WAV file?
      - Can I export just a time slice of an audio sample?
      instructions:
      - text: Get sample {sampleId} in project {projectId} as WAV using axis {axisIx}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
          axisIx: query.axisIx
      - text: Export axis {axisIx} of sample {sampleId} in project {projectId} as WAV from {sliceStart} to {sliceEnd}.
        slots:
          axisIx: query.axisIx
          sampleId: path.sampleId
          projectId: path.projectId
          sliceStart: query.sliceStart
          sliceEnd: query.sliceEnd
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/{sampleId}/image'].get
  update:
    x-apievangelist-phrasing:
      intent: Get an image sample as an image file
      effect: read
      questions:
      - How do I view an RGBA sample as an image file?
      - Can I see what an image looks like after the impulse's input block processes it?
      instructions:
      - text: Get sample {sampleId} in project {projectId} as an image file.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
      - text: Render image sample {sampleId} from project {projectId} after the input block of impulse {impulseId}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
          impulseId: query.impulseId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/{sampleId}/video'].get
  update:
    x-apievangelist-phrasing:
      intent: Get a video sample as a video file
      effect: read
      questions:
      - How do I play back a video sample from my dataset?
      - Can I download a sample containing video data as a video file?
      instructions:
      - text: Get sample {sampleId} in project {projectId} as a video file.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
      - text: Download the video for sample {sampleId} of project {projectId}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/{sampleId}/slice'].get
  update:
    x-apievangelist-phrasing:
      intent: Get a slice of a sample's data
      effect: read
      questions:
      - Can I fetch only part of a long sample's data between two points?
      - How do I read a window of values from a sample starting at a given offset?
      instructions:
      - text: Get a slice of sample {sampleId} in project {projectId} starting at {sliceStart}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
          sliceStart: query.sliceStart
      - text: Return sample {sampleId} data from {sliceStart} to {sliceEnd} in project {projectId}.
        slots:
          sampleId: path.sampleId
          sliceStart: query.sliceStart
          sliceEnd: query.sliceEnd
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/{sampleId}/rename'].post
  update:
    x-apievangelist-phrasing:
      intent: Rename a sample
      effect: write
      questions:
      - How do I change the file name of a sample?
      - Do sample names need to be unique within a project?
      instructions:
      - text: Rename sample {sampleId} in project {projectId} to {name}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
          name: requestBody.name
      - text: Set the file name of sample {sampleId} in project {projectId} to {name}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
          name: requestBody.name
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/{sampleId}/edit-label'].post
  update:
    x-apievangelist-phrasing:
      intent: Change the label of one sample
      effect: write
      questions:
      - How do I fix the class label on a single mislabeled sample?
      - Can I relabel one sample without touching the rest of the dataset?
      instructions:
      - text: Set the label of sample {sampleId} in project {projectId} to {label}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
          label: requestBody.label
      - text: Relabel just sample {sampleId} of project {projectId} as {label}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
          label: requestBody.label
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/{sampleId}/move'].post
  update:
    x-apievangelist-phrasing:
      intent: Move one sample to another category
      effect: write
      questions:
      - How do I move a single sample from the test set into training?
      - Can I reassign one sample's dataset category?
      instructions:
      - text: Move sample {sampleId} in project {projectId} to {newCategory}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
          newCategory: requestBody.newCategory
      - text: Put the single sample {sampleId} of project {projectId} into the {newCategory} set.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
          newCategory: requestBody.newCategory
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/{sampleId}/crop'].post
  update:
    x-apievangelist-phrasing:
      intent: Crop a sample to a new range
      effect: write
      questions:
      - How do I trim the silence off the start and end of a recording?
      - Can I crop a sample to keep only a section between two points?
      instructions:
      - text: Crop sample {sampleId} in project {projectId} from {cropStart} to {cropEnd}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
          cropStart: requestBody.cropStart
          cropEnd: requestBody.cropEnd
      - text: Trim sample {sampleId} of project {projectId} to the range {cropStart} through {cropEnd}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
          cropStart: requestBody.cropStart
          cropEnd: requestBody.cropEnd
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/{sampleId}/split'].post
  update:
    x-apievangelist-phrasing:
      intent: Split a video sample into frames
      effect: write
      questions:
      - How do I turn a video sample into individual image frames?
      - Can I choose the frame rate when splitting a video into frames?
      instructions:
      - text: Split video sample {sampleId} in project {projectId} into frames.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
      - text: Extract frames from video {sampleId} of project {projectId} at {fps} fps.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
          fps: requestBody.fps
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/{sampleId}/find-segments'].post
  update:
    x-apievangelist-phrasing:
      intent: Detect non-noise events in a sample
      effect: read
      questions:
      - Can Edge Impulse automatically find where the interesting events are in a long recording?
      - How do I get start and end times for every non-noise event in a sample?
      instructions:
      - text: Find segments in sample {sampleId} of project {projectId} with {segmentLengthMs} ms segments.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
          segmentLengthMs: requestBody.segmentLengthMs
      - text: Detect events in sample {sampleId} in project {projectId}, {segmentLengthMs} ms long, shifting segments {shiftSegments}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
          segmentLengthMs: requestBody.segmentLengthMs
          shiftSegments: requestBody.shiftSegments
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/{sampleId}/segment'].post
  update:
    x-apievangelist-phrasing:
      intent: Slice a sample into multiple segments
      effect: write
      questions:
      - How do I cut one long sample into several shorter samples?
      - What happens to the original file after I segment a sample?
      instructions:
      - text: Segment sample {sampleId} in project {projectId} into {segments}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
          segments: requestBody.segments
      - text: 'Slice sample {sampleId} of project {projectId} at these segment boundaries: {segments}.'
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
          segments: requestBody.segments
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/{sampleId}/bounding-boxes'].post
  update:
    x-apievangelist-phrasing:
      intent: Set bounding boxes on one sample
      effect: write
      questions:
      - How do I save bounding-box annotations for a single image?
      - Can I replace the object boxes drawn on one sample?
      instructions:
      - text: Set the bounding boxes of sample {sampleId} in project {projectId} to {boundingBoxes}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
          boundingBoxes: requestBody.boundingBoxes
      - text: Annotate image {sampleId} of project {projectId} with boxes {boundingBoxes}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
          boundingBoxes: requestBody.boundingBoxes
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/{sampleId}/structured-labels'].post
  update:
    x-apievangelist-phrasing:
      intent: Set multiple time-ranged labels on a sample
      effect: write
      questions:
      - Can one sample carry several labels over different time ranges?
      - How do I label parts of a sample with start and end indexes?
      instructions:
      - text: Set structured labels {structuredLabels} on sample {sampleId} in project {projectId}.
        slots:
          structuredLabels: requestBody.structuredLabels
          sampleId: path.sampleId
          projectId: path.projectId
      - text: Give sample {sampleId} of project {projectId} multi-label ranges {structuredLabels}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
          structuredLabels: requestBody.structuredLabels
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/store-segment-length'].post
  update:
    x-apievangelist-phrasing:
      intent: Remember the last segment length used
      effect: write
      questions:
      - Can the project remember my segment length so all segments come out uniform?
      - How do I save a default segment length for future segmenting?
      instructions:
      - text: Store a segment length of {segmentLength} for project {projectId}.
        slots:
          segmentLength: requestBody.segmentLength
          projectId: path.projectId
      - text: Remember {segmentLength} as the last segment length in project {projectId}.
        slots:
          segmentLength: requestBody.segmentLength
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/{sampleId}/original'].get
  update:
    x-apievangelist-phrasing:
      intent: Get a sample's original uncropped data
      effect: read
      questions:
      - Can I see a sample's data as it was before I cropped it?
      - How do I zoom into a range of the original downsampled signal?
      instructions:
      - text: Get the original uncropped data for sample {sampleId} in project {projectId}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
      - text: Show uncropped downsampled data of sample {sampleId} in project {projectId} zoomed from {zoomStart} to {zoomEnd}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
          zoomStart: query.zoomStart
          zoomEnd: query.zoomEnd
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/{sampleId}/retry-processing'].post
  update:
    x-apievangelist-phrasing:
      intent: Retry processing a failed sample
      effect: write
      questions:
      - What can I do about a sample that failed processing after upload?
      - Is there a way to reprocess a single broken sample?
      instructions:
      - text: Retry processing sample {sampleId} in project {projectId}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
      - text: Reprocess failed sample {sampleId} of project {projectId}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/{sampleId}/enable'].post
  update:
    x-apievangelist-phrasing:
      intent: Re-include one sample in the dataset
      effect: write
      questions:
      - How do I bring a disabled sample back into my dataset?
      - Can I re-enable a single excluded sample?
      instructions:
      - text: Enable sample {sampleId} in project {projectId}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
      - text: Include the single sample {sampleId} of project {projectId} in the dataset again.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/{sampleId}/disable'].post
  update:
    x-apievangelist-phrasing:
      intent: Exclude one sample from the dataset
      effect: write
      questions:
      - Can I exclude one sample from training without deleting it?
      - How do I temporarily disable a single questionable sample?
      instructions:
      - text: Disable sample {sampleId} in project {projectId}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
      - text: Exclude the single sample {sampleId} of project {projectId} from the dataset.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/{sampleId}/autolabel'].post
  update:
    x-apievangelist-phrasing:
      intent: Auto-label an image with another neural network
      effect: write
      questions:
      - Can another neural network classify an image for me automatically?
      - How do I auto-label one image sample?
      instructions:
      - text: Auto-label image {sampleId} in project {projectId} using {neuralNetwork}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
          neuralNetwork: requestBody.neuralNetwork
      - text: Classify sample {sampleId} of project {projectId} with neural network {neuralNetwork}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
          neuralNetwork: requestBody.neuralNetwork
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/{sampleId}/metadata'].post
  update:
    x-apievangelist-phrasing:
      intent: Set metadata on one sample
      effect: write
      questions:
      - How do I attach metadata such as device or location to a single sample?
      - Can I update the key-value metadata on one sample?
      instructions:
      - text: Set metadata {metadata} on sample {sampleId} in project {projectId}.
        slots:
          metadata: requestBody.metadata
          sampleId: path.sampleId
          projectId: path.projectId
      - text: Update the single sample {sampleId} of project {projectId} with metadata {metadata}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
          metadata: requestBody.metadata
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/{sampleId}/to-labeling-queue'].post
  update:
    x-apievangelist-phrasing:
      intent: Send one sample back to the labeling queue
      effect: destructive
      questions:
      - How do I redo the bounding boxes on a single image?
      - Can I clear one image's boxes and put it back in the labeling queue?
      instructions:
      - text: Move sample {sampleId} in project {projectId} back to the labeling queue.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
      - text: Clear the boxes on image {sampleId} of project {projectId} and requeue it for labeling.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/{sampleId}/propose-changes'].post
  update:
    x-apievangelist-phrasing:
      intent: Queue AI Actions proposed changes for a sample
      effect: write
      questions:
      - How do I stage changes to a sample as part of an AI Actions job?
      - Will proposing new changes overwrite earlier proposals for the same sample?
      instructions:
      - text: Propose changes {proposedChanges} to sample {sampleId} in project {projectId} for AI Actions job {jobId}.
        slots:
          proposedChanges: requestBody.proposedChanges
          sampleId: path.sampleId
          projectId: path.projectId
          jobId: requestBody.jobId
      - text: Queue job {jobId} proposals {proposedChanges} on sample {sampleId} of project {projectId}.
        slots:
          jobId: requestBody.jobId
          proposedChanges: requestBody.proposedChanges
          sampleId: path.sampleId
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/batch/edit-labels'].post
  update:
    x-apievangelist-phrasing:
      intent: Relabel many samples at once
      effect: write
      questions:
      - How do I change the label on a whole batch of samples in one call?
      - Can I relabel every training sample whose filename matches a pattern?
      instructions:
      - text: Relabel all {category} samples currently labeled {labels} in project {projectId} as {label}.
        slots:
          category: query.category
          labels: query.labels
          projectId: path.projectId
          label: requestBody.label
      - text: Set label {label} on {category} samples {ids} in project {projectId}.
        slots:
          label: requestBody.label
          category: query.category
          ids: query.ids
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/batch/delete'].post
  update:
    x-apievangelist-phrasing:
      intent: Delete many samples matching a filter
      effect: destructive
      questions:
      - How do I delete every sample with a particular label in bulk?
      - Can I remove a specific list of sample IDs in one request?
      instructions:
      - text: Delete all {category} samples labeled {labels} in project {projectId}.
        slots:
          category: query.category
          labels: query.labels
          projectId: path.projectId
      - text: Bulk delete {category} samples {ids} from project {projectId}.
        slots:
          category: query.category
          ids: query.ids
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/batch/moveSamples'].post
  update:
    x-apievangelist-phrasing:
      intent: Move many samples to another category
      effect: write
      questions:
      - How do I move a batch of test samples into the training set?
      - Can I shift all samples of one label to a different category at once?
      instructions:
      - text: Move all {category} samples labeled {labels} in project {projectId} to {newCategory}.
        slots:
          category: query.category
          labels: query.labels
          projectId: path.projectId
          newCategory: requestBody.newCategory
      - text: Bulk move {category} samples {ids} of project {projectId} into {newCategory}.
        slots:
          category: query.category
          ids: query.ids
          projectId: path.projectId
          newCategory: requestBody.newCategory
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/batch/enable-samples'].post
  update:
    x-apievangelist-phrasing:
      intent: Re-include many samples in the dataset
      effect: write
      questions:
      - How do I re-enable a whole group of disabled samples?
      - Can I bring back every excluded sample with a given label?
      instructions:
      - text: Enable all {category} samples labeled {labels} in project {projectId}.
        slots:
          category: query.category
          labels: query.labels
          projectId: path.projectId
      - text: Bulk enable {category} samples {ids} in project {projectId}.
        slots:
          category: query.category
          ids: query.ids
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/batch/disable-samples'].post
  update:
    x-apievangelist-phrasing:
      intent: Exclude many samples from the dataset
      effect: write
      questions:
      - How do I exclude a batch of samples from training without deleting them?
      - Can I disable every sample shorter than a minimum length?
      instructions:
      - text: Disable all {category} samples labeled {labels} in project {projectId}.
        slots:
          category: query.category
          labels: query.labels
          projectId: path.projectId
      - text: Bulk disable {category} samples shorter than {minLength} in project {projectId}.
        slots:
          category: query.category
          minLength: query.minLength
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/batch/add-metadata'].post
  update:
    x-apievangelist-phrasing:
      intent: Add a metadata field to many samples
      effect: write
      questions:
      - How do I tag a batch of samples with the same metadata key and value?
      - Can I add metadata to every sample with a certain label in one step?
      instructions:
      - text: Add metadata {metadataKey}={metadataValue} to all {category} samples in project {projectId}.
        slots:
          metadataKey: requestBody.metadataKey
          metadataValue: requestBody.metadataValue
          category: query.category
          projectId: path.projectId
      - text: Tag {category} samples labeled {labels} in project {projectId} with {metadataKey} set to {metadataValue}.
        slots:
          category: query.category
          labels: query.labels
          projectId: path.projectId
          metadataKey: requestBody.metadataKey
          metadataValue: requestBody.metadataValue
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/batch/clear-metadata-by-key'].post
  update:
    x-apievangelist-phrasing:
      intent: Remove one metadata key from many samples
      effect: destructive
      questions:
      - How do I strip a single metadata field from a batch of samples?
      - Can I clear one metadata key across all my training samples?
      instructions:
      - text: Clear metadata key {metadataKey} from all {category} samples in project {projectId}.
        slots:
          metadataKey: requestBody.metadataKey
          category: query.category
          projectId: path.projectId
      - text: Remove the {metadataKey} field from {category} samples labeled {labels} in project {projectId}.
        slots:
          metadataKey: requestBody.metadataKey
          category: query.category
          labels: query.labels
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/batch/clear-metadata'].post
  update:
    x-apievangelist-phrasing:
      intent: Clear all metadata from many samples
      effect: destructive
      questions:
      - How do I wipe every metadata field from a set of samples?
      - Can I reset all metadata on samples that match a filter?
      instructions:
      - text: Wipe every metadata key and value on the {category} set of project {projectId}.
        slots:
          category: query.category
          projectId: path.projectId
      - text: Remove every metadata field from {category} samples {ids} in project {projectId}.
        slots:
          category: query.category
          ids: query.ids
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/track-objects'].post
  update:
    x-apievangelist-phrasing:
      intent: Track labeled objects into the next sample
      effect: write
      questions:
      - Can bounding boxes from one frame be carried forward to the next frame automatically?
      - How do I track objects between two consecutive image samples?
      instructions:
      - text: Track objects from sample {sourceSampleId} into sample {nextSampleId} in project {projectId}.
        slots:
          sourceSampleId: requestBody.sourceSampleId
          nextSampleId: requestBody.nextSampleId
          projectId: path.projectId
      - text: Carry bounding boxes on {sourceSampleId} forward to {nextSampleId} in project {projectId}.
        slots:
          sourceSampleId: requestBody.sourceSampleId
          nextSampleId: requestBody.nextSampleId
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/data-explorer/features'].get
  update:
    x-apievangelist-phrasing:
      intent: Get the data explorer's 2D feature map
      effect: read
      questions:
      - How can I get the t-SNE projection that the data explorer plots?
      - Where do I fetch the 2D coordinates of every sample in the data explorer?
      instructions:
      - text: Get the data explorer features for project {projectId}.
        slots:
          projectId: path.projectId
      - text: Return the t-SNE points plotted in the data explorer of project {projectId}.
        slots:
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/data-explorer/has-features'].get
  update:
    x-apievangelist-phrasing:
      intent: Check if data explorer features exist
      effect: read
      questions:
      - Has the data explorer already been generated for my project?
      - Is there a computed feature map ready before I try to load the explorer?
      instructions:
      - text: Check whether project {projectId} has data explorer features.
        slots:
          projectId: path.projectId
      - text: Tell me if the data explorer for project {projectId} has been computed.
        slots:
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/data-explorer/predictions'].get
  update:
    x-apievangelist-phrasing:
      intent: Get model predictions for data explorer points
      effect: read
      questions:
      - What did the current impulse predict for each point in the data explorer?
      - Can I see which data explorer samples the model gets wrong?
      instructions:
      - text: Get data explorer predictions for project {projectId}.
        slots:
          projectId: path.projectId
      - text: Show the impulse's prediction for every explorer point in project {projectId}.
        slots:
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/data-explorer/clear'].post
  update:
    x-apievangelist-phrasing:
      intent: Clear the data explorer state
      effect: destructive
      questions:
      - How do I reset the data explorer so it can be regenerated?
      - Can I throw away the current data explorer results?
      instructions:
      - text: Clear the data explorer for project {projectId}.
        slots:
          projectId: path.projectId
      - text: Remove the current data explorer state in project {projectId}.
        slots:
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/data-explorer/settings'].get
  update:
    x-apievangelist-phrasing:
      intent: Get data explorer settings
      effect: read
      questions:
      - Which preset and dimensionality reduction does my data explorer use?
      - What input and DSP block is the data explorer configured with?
      instructions:
      - text: Get the data explorer settings for project {projectId}.
        slots:
          projectId: path.projectId
      - text: Show how the data explorer is configured in project {projectId}.
        slots:
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/data-explorer/settings'].post
  update:
    x-apievangelist-phrasing:
      intent: Configure the data explorer
      effect: write
      questions:
      - How do I switch the data explorer to a different dimensionality reduction technique?
      - Can I point the data explorer at a specific impulse?
      instructions:
      - text: Set the data explorer in project {projectId} to use preset {preset}.
        slots:
          projectId: path.projectId
          preset: requestBody.preset
      - text: Configure project {projectId} data explorer with {dimensionalityReductionTechnique} on impulse {impulseId}.
        slots:
          projectId: path.projectId
          dimensionalityReductionTechnique: requestBody.dimensionalityReductionTechnique
          impulseId: requestBody.impulseId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/data-explorer/screenshot'].post
  update:
    x-apievangelist-phrasing:
      intent: Upload a data explorer screenshot
      effect: write
      questions:
      - Can a data pipeline upload a picture of the data explorer?
      - Where does a screenshot of the data explorer get stored from a pipeline?
      instructions:
      - text: Upload data explorer screenshot {image} to project {projectId}.
        slots:
          image: requestBody.image
          projectId: path.projectId
      - text: Save image {image} as the data explorer snapshot for project {projectId}.
        slots:
          image: requestBody.image
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/data-quality/diversity/exists'].get
  update:
    x-apievangelist-phrasing:
      intent: Check if diversity metrics were calculated
      effect: read
      questions:
      - Have dataset diversity metrics been calculated for my project yet?
      - Is there similarity and diversity data available to load?
      instructions:
      - text: Check whether diversity metrics exist for project {projectId}.
        slots:
          projectId: path.projectId
      - text: Tell me if project {projectId} has calculated data diversity metrics.
        slots:
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/data-quality/diversity'].get
  update:
    x-apievangelist-phrasing:
      intent: Get dataset diversity metrics
      effect: read
      questions:
      - How diverse or repetitive are the samples in my dataset?
      - Which samples look too similar to each other?
      instructions:
      - text: Get the diversity metrics for project {projectId}.
        slots:
          projectId: path.projectId
      - text: Show similarity and diversity data for the dataset of project {projectId}.
        slots:
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/data-quality/label-noise/exists'].get
  update:
    x-apievangelist-phrasing:
      intent: Check if label noise metrics were calculated
      effect: read
      questions:
      - Have label noise metrics been computed for my dataset?
      - Is a mislabeling analysis available for my project yet?
      instructions:
      - text: Check whether label noise metrics exist for project {projectId}.
        slots:
          projectId: path.projectId
      - text: Tell me if project {projectId} has computed label noise data.
        slots:
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/data-quality/label-noise'].get
  update:
    x-apievangelist-phrasing:
      intent: Get label noise metrics
      effect: read
      questions:
      - Which samples in my dataset are likely mislabeled?
      - Can I see potential label noise issues across the dataset?
      instructions:
      - text: Get label noise data for project {projectId}.
        slots:
          projectId: path.projectId
      - text: Show possibly mislabeled samples in project {projectId}.
        slots:
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/ai-actions-preview/{jobId}/proposed-changes'].get
  update:
    x-apievangelist-phrasing:
      intent: Get proposed changes from an AI Actions job
      effect: read
      questions:
      - What changes is an AI Actions job proposing to my samples?
      - Can I preview AI Actions results before they are applied?
      instructions:
      - text: Get the proposed changes from AI Actions job {jobId} in project {projectId}.
        slots:
          jobId: path.jobId
          projectId: path.projectId
      - text: Preview what AI Actions job {jobId} would change in project {projectId}.
        slots:
          jobId: path.jobId
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/v2/rebalance'].post
  update:
    x-apievangelist-phrasing:
      intent: Rebalance train/test split (v2 endpoint)
      effect: write
      questions:
      - Is there a v2 version of the rebalance call for resetting to an 80/20 split?
      - How do I redistribute samples between training and testing with the newer rebalance endpoint?
      instructions:
      - text: Rebalance project {projectId} to 80/20 train/test using the v2 rebalance endpoint.
        slots:
          projectId: path.projectId
      - text: Run the v2 dataset rebalance on project {projectId}.
        slots:
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/split'].post
  update:
    x-apievangelist-phrasing:
      intent: Split the dataset into train, test and validation
      effect: write
      questions:
      - How do I split my dataset with custom ratios including a validation set?
      - Can I keep samples sharing a metadata key together in the same split?
      - Is it possible to stratify the train/test split by label?
      instructions:
      - text: Split project {projectId} into {trainingSplitRatio} training and {testingSplitRatio} testing.
        slots:
          projectId: path.projectId
          trainingSplitRatio: requestBody.trainingSplitRatio
          testingSplitRatio: requestBody.testingSplitRatio
      - text: Apply a {trainingSplitRatio}/{testingSplitRatio}/{validationSplitRatio} split to project {projectId}, stratified by {stratifyBy}.
        slots:
          trainingSplitRatio: requestBody.trainingSplitRatio
          testingSplitRatio: requestBody.testingSplitRatio
          validationSplitRatio: requestBody.validationSplitRatio
          projectId: path.projectId
          stratifyBy: requestBody.stratifyBy
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/split-preview'].post
  update:
    x-apievangelist-phrasing:
      intent: Preview a dataset split before applying it
      effect: read
      questions:
      - Can I see how my samples would be distributed before actually splitting?
      - What would each label's share look like under a proposed split ratio?
      instructions:
      - text: Preview a {trainingSplitRatio}/{testingSplitRatio} split for project {projectId} without applying it.
        slots:
          trainingSplitRatio: requestBody.trainingSplitRatio
          testingSplitRatio: requestBody.testingSplitRatio
          projectId: path.projectId
      - text: Start a split preview for project {projectId} at {trainingSplitRatio} train, {testingSplitRatio} test, keeping {keepTogetherMetadataKeys} together.
        slots:
          projectId: path.projectId
          trainingSplitRatio: requestBody.trainingSplitRatio
          testingSplitRatio: requestBody.testingSplitRatio
          keepTogetherMetadataKeys: requestBody.keepTogetherMetadataKeys
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/get-split-preview-results'].get
  update:
    x-apievangelist-phrasing:
      intent: Get results of a split preview job
      effect: read
      questions:
      - Where do I collect the results of the split preview I started earlier?
      - Has my split preview job finished, and what did it show?
      instructions:
      - text: Get the split preview results for project {projectId}.
        slots:
          projectId: path.projectId
      - text: Fetch the finished split preview output of project {projectId}.
        slots:
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/ratio'].get
  update:
    x-apievangelist-phrasing:
      intent: Get the train/test sample counts
      effect: read
      questions:
      - What is the current ratio of training to testing samples in my project?
      - How many samples sit in each of train and test right now?
      instructions:
      - text: Get the train/test ratio for project {projectId}.
        slots:
          projectId: path.projectId
      - text: Show how many samples are in training versus testing for project {projectId}.
        slots:
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/metadata-filter-options'].get
  update:
    x-apievangelist-phrasing:
      intent: List metadata values usable as sample filters
      effect: read
      questions:
      - Which metadata key-value pairs can I filter my samples by?
      - What unique metadata values exist across the training set?
      instructions:
      - text: List metadata filter options for {category} samples in project {projectId}.
        slots:
          category: query.category
          projectId: path.projectId
      - text: Show the distinct metadata key/value pairs across {category} data in project {projectId}.
        slots:
          category: query.category
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/project-metadata'].get
  update:
    x-apievangelist-phrasing:
      intent: Get dataset-level metadata for a project
      effect: read
      questions:
      - What overall dataset metadata does my project have?
      - Can I get dataset metadata for a specific dataset version?
      instructions:
      - text: Get the dataset metadata for project {projectId}.
        slots:
          projectId: path.projectId
      - text: Show dataset metadata of project {projectId} for version {datasetVersionId}.
        slots:
          projectId: path.projectId
          datasetVersionId: query.datasetVersionId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/{sampleId}/video-dimensions'].post
  update:
    x-apievangelist-phrasing:
      intent: Set a video sample's dimensions
      effect: write
      questions:
      - How do I set the width and height on a video sample that has none?
      - Can I fix missing dimensions on an uploaded video?
      instructions:
      - text: Set video sample {sampleId} in project {projectId} to {width}x{height}.
        slots:
          sampleId: path.sampleId
          projectId: path.projectId
          width: requestBody.width
          height: requestBody.height
      - text: Record dimensions width {width} and height {height} for video {sampleId} of project {projectId}.
        slots:
          width: requestBody.width
          height: requestBody.height
          sampleId: path.sampleId
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/batch/edit-bounding-boxes'].post
  update:
    x-apievangelist-phrasing:
      intent: Relabel or remove bounding boxes in bulk
      effect: write
      questions:
      - How do I rename a bounding-box class across many images at once?
      - Can I merge several box labels into one across the training set?
      instructions:
      - text: Relabel bounding boxes {oldLabels} as {newLabel} across {category} samples in project {projectId}.
        slots:
          oldLabels: requestBody.oldLabels
          newLabel: requestBody.newLabel
          category: query.category
          projectId: path.projectId
      - text: Change box labels {oldLabels} to {newLabel} on {category} samples {ids} of project {projectId}.
        slots:
          oldLabels: requestBody.oldLabels
          newLabel: requestBody.newLabel
          category: query.category
          ids: query.ids
          projectId: path.projectId
      method: generated
      generated: '2026-10-01'
- target: $.paths['/api/{projectId}/raw-data/batch/back-to-labeling'].post
  update:
    x-apievangelist-phrasing:
      intent: Send many samples back to the labeling queue
      effect: write
      questions:
      - How do I put a batch of images back into the object detection labeling queue?
      - Can I requeue every image with a certain label for relabeling?
      instructions:
      - text: Put {category} samples labeled {labels} in project {projectId} back into labeling with action {labelAction}.
        slots:
          category: query.category
          labels: query.labels
          projectId: path.projectId
          labelAction: requestBody.labelAction
      - text: Bulk requeue {category} samples {ids} of project {projectId} for labeling, action {labelAction}.
        slots:
          category: query.category
          ids: query.ids
          projectId: path.projectId
          labelAction: requestBody.labelAction
      method: generated
      generated: '2026-10-01'