# Generated by API Evangelist (build-phrasing.py). Our phrasing, not observed demand.
overlay: 1.0.0
info:
  title: API Evangelist conversational phrasing for Elasticsearch Request & Response Specification ml data frame API
  version: 1.0.0
extends: openapi/elk-stack-ml-data-frame-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: 18
- target: $.paths['/_ml/data_frame/analytics/{id}'].get
  update:
    x-apievangelist-phrasing:
      intent: Get data frame analytics job configs by ID
      effect: read
      questions:
      - What source, destination and analysis is a specific data frame analytics job configured with?
      - Can I fetch several data frame analytics jobs at once with a comma-separated list or wildcard?
      instructions:
      - text: Show the configuration of data frame analytics job {id}.
        slots:
          id: path.id
      - text: Get configs for the data frame analytics jobs matching {id}, excluding generated fields.
        slots:
          id: path.id
      method: generated
      generated: '2026-09-26'
- target: $.paths['/_ml/data_frame/analytics/{id}'].put
  update:
    x-apievangelist-phrasing:
      intent: Create a data frame analytics job
      effect: write
      questions:
      - How do I set up outlier detection, regression or classification on an index?
      - Can I cap the memory a new data frame analytics job is allowed to use?
      - Will the destination index be created for me when the analytics job starts?
      instructions:
      - text: Create data frame analytics job {id} reading from {source}, writing to {dest}, running {analysis}.
        slots:
          id: path.id
          source: requestBody.source
          dest: requestBody.dest
          analysis: requestBody.analysis
      - text: Create analytics job {id} with a model memory limit of {model_memory_limit}.
        slots:
          id: path.id
          model_memory_limit: requestBody.model_memory_limit
      method: generated
      generated: '2026-09-26'
- target: $.paths['/_ml/data_frame/analytics/{id}'].delete
  update:
    x-apievangelist-phrasing:
      intent: Delete a data frame analytics job
      effect: destructive
      questions:
      - How do I remove a data frame analytics job I no longer need?
      - Can I force-delete a data frame analytics job that is still running?
      instructions:
      - text: Delete data frame analytics job {id}.
        slots:
          id: path.id
      - text: Force delete the analytics job {id} even though it is running.
        slots:
          id: path.id
      method: generated
      generated: '2026-09-26'
- target: $.paths['/_ml/data_frame/_evaluate'].post
  update:
    x-apievangelist-phrasing:
      intent: Evaluate data frame analytics results
      effect: read
      questions:
      - How accurate is my classification model compared with the ground truth in its results index?
      - Can I compute evaluation metrics like AUC ROC or mean squared error on an analytics results index?
      instructions:
      - text: Evaluate the analytics results in index {index} using {evaluation}.
        slots:
          index: requestBody.index
          evaluation: requestBody.evaluation
      - text: Compute evaluation metrics {evaluation} for index {index}, limited to documents matching {query}.
        slots:
          evaluation: requestBody.evaluation
          index: requestBody.index
          query: requestBody.query
      method: generated
      generated: '2026-09-26'
- target: $.paths['/_ml/data_frame/analytics/_explain'].get
  update:
    x-apievangelist-phrasing:
      intent: Explain an unsaved analytics config via GET
      effect: read
      questions:
      - Before creating a job, which fields would a draft data frame analytics config include or skip, using the GET form?
      - What memory would a not-yet-created analytics job need, asked over GET?
      instructions:
      - text: Explain field selection and memory for a draft analytics config on {source} via GET.
        slots:
          source: requestBody.source
      - text: Using the GET explain call, estimate memory for analysis {analysis} before I create the job.
        slots:
          analysis: requestBody.analysis
      method: generated
      generated: '2026-09-26'
- target: $.paths['/_ml/data_frame/analytics/_explain'].post
  update:
    x-apievangelist-phrasing:
      intent: Explain an unsaved analytics config via POST
      effect: read
      questions:
      - Can I POST a draft analytics config and get back which fields it would analyze?
      - How much memory would a proposed analytics job need, sending the draft config by POST?
      instructions:
      - text: POST a draft analytics config with source {source} and dest {dest} and explain its field selection.
        slots:
          source: requestBody.source
          dest: requestBody.dest
      - text: Send analysis {analysis} by POST to estimate its memory before creating a job.
        slots:
          analysis: requestBody.analysis
      method: generated
      generated: '2026-09-26'
- target: $.paths['/_ml/data_frame/analytics/{id}/_explain'].get
  update:
    x-apievangelist-phrasing:
      intent: Explain an existing analytics job via GET
      effect: read
      questions:
      - Why were certain fields left out of an existing data frame analytics job, checked with GET?
      - What memory does an already-created analytics job estimate it needs, via the GET explain?
      instructions:
      - text: Explain the field selection of existing analytics job {id} using GET.
        slots:
          id: path.id
      - text: GET the memory estimate for data frame analytics job {id}.
        slots:
          id: path.id
      method: generated
      generated: '2026-09-26'
- target: $.paths['/_ml/data_frame/analytics/{id}/_explain'].post
  update:
    x-apievangelist-phrasing:
      intent: Explain an existing analytics job via POST
      effect: read
      questions:
      - Can I POST to explain an existing analytics job, overriding some of its settings in the body?
      - How would changing analyzed_fields affect an existing job's memory estimate, sent by POST?
      instructions:
      - text: POST an explain for analytics job {id} with analyzed fields {analyzed_fields}.
        slots:
          id: path.id
          analyzed_fields: requestBody.analyzed_fields
      - text: Explain existing job {id} by POST and report the estimated memory.
        slots:
          id: path.id
      method: generated
      generated: '2026-09-26'
- target: $.paths['/_ml/data_frame/analytics'].get
  update:
    x-apievangelist-phrasing:
      intent: List all data frame analytics jobs
      effect: read
      questions:
      - What data frame analytics jobs exist on my cluster?
      - Can I page through all analytics job configurations using from and size?
      instructions:
      - text: List every data frame analytics job configuration.
      - text: Show {size} data frame analytics jobs starting at {from}.
        slots:
          size: query.size
          from: query.from
      method: generated
      generated: '2026-09-26'
- target: $.paths['/_ml/data_frame/analytics/_stats'].get
  update:
    x-apievangelist-phrasing:
      intent: Get stats for all data frame analytics jobs
      effect: read
      questions:
      - What is the progress and state of all my data frame analytics jobs?
      - Can I get verbose statistics across every analytics job at once?
      instructions:
      - text: Show usage stats and progress for all data frame analytics jobs.
      - text: Get verbose stats for every analytics job, {size} at a time.
        slots:
          size: query.size
      method: generated
      generated: '2026-09-26'
- target: $.paths['/_ml/data_frame/analytics/{id}/_stats'].get
  update:
    x-apievangelist-phrasing:
      intent: Get stats for a data frame analytics job
      effect: read
      questions:
      - How far along is a particular data frame analytics job?
      - Is a specific analytics job running, stopped or failed?
      instructions:
      - text: Show the progress and state of analytics job {id}.
        slots:
          id: path.id
      - text: Get verbose stats for data frame analytics job {id}.
        slots:
          id: path.id
      method: generated
      generated: '2026-09-26'
- target: $.paths['/_ml/data_frame/analytics/_preview'].get
  update:
    x-apievangelist-phrasing:
      intent: Preview features for a draft config via GET
      effect: read
      questions:
      - What extracted feature values would a draft analytics config produce, previewed with GET?
      - Can I sample the training features of a config I haven't saved yet using a GET request?
      instructions:
      - text: Preview the extracted features for draft config {config} via GET.
        slots:
          config: requestBody.config
      - text: GET a feature preview for this unsaved analytics config {config}.
        slots:
          config: requestBody.config
      method: generated
      generated: '2026-09-26'
- target: $.paths['/_ml/data_frame/analytics/_preview'].post
  update:
    x-apievangelist-phrasing:
      intent: Preview features for a draft config via POST
      effect: read
      questions:
      - Can I POST a draft analytics config and see sample extracted features?
      - What would the feature rows look like for a proposed config sent by POST?
      instructions:
      - text: POST draft config {config} and show me the extracted feature preview.
        slots:
          config: requestBody.config
      - text: Preview features of proposed analytics config {config} with a POST request.
        slots:
          config: requestBody.config
      method: generated
      generated: '2026-09-26'
- target: $.paths['/_ml/data_frame/analytics/{id}/_preview'].get
  update:
    x-apievangelist-phrasing:
      intent: Preview features of an existing job via GET
      effect: read
      questions:
      - What feature values does an existing analytics job extract, previewed via GET?
      - Can I check the training data an existing job will see before starting it, using GET?
      instructions:
      - text: GET a feature preview for analytics job {id}.
        slots:
          id: path.id
      - text: Show sample extracted features for existing job {id} via GET.
        slots:
          id: path.id
      method: generated
      generated: '2026-09-26'
- target: $.paths['/_ml/data_frame/analytics/{id}/_preview'].post
  update:
    x-apievangelist-phrasing:
      intent: Preview features of an existing job via POST
      effect: read
      questions:
      - Can I preview an existing analytics job's features with a POST request?
      - What rows would an existing job extract, fetched by POST?
      instructions:
      - text: POST a feature preview request for analytics job {id}.
        slots:
          id: path.id
      - text: Preview extracted features for existing job {id} using POST.
        slots:
          id: path.id
      method: generated
      generated: '2026-09-26'
- target: $.paths['/_ml/data_frame/analytics/{id}/_start'].post
  update:
    x-apievangelist-phrasing:
      intent: Start a data frame analytics job
      effect: write
      questions:
      - How do I kick off a data frame analytics job I've created?
      - Can I restart an analytics job after stopping it?
      instructions:
      - text: Start data frame analytics job {id}.
        slots:
          id: path.id
      - text: Start analytics job {id} and wait up to {timeout} for it to begin.
        slots:
          id: path.id
          timeout: query.timeout
      method: generated
      generated: '2026-09-26'
- target: $.paths['/_ml/data_frame/analytics/{id}/_stop'].post
  update:
    x-apievangelist-phrasing:
      intent: Stop data frame analytics jobs
      effect: write
      questions:
      - How can I pause a running data frame analytics job?
      - Is there a force option to stop an analytics job that won't stop cleanly?
      instructions:
      - text: Stop data frame analytics job {id}.
        slots:
          id: path.id
      - text: Force stop analytics job {id} within {timeout}.
        slots:
          id: path.id
          timeout: query.timeout
      method: generated
      generated: '2026-09-26'
- target: $.paths['/_ml/data_frame/analytics/{id}/_update'].post
  update:
    x-apievangelist-phrasing:
      intent: Update a data frame analytics job
      effect: write
      questions:
      - Can I raise the memory limit on an analytics job after creating it?
      - Which settings of a data frame analytics job can I change, like description or max threads?
      instructions:
      - text: Update analytics job {id} to a model memory limit of {model_memory_limit}.
        slots:
          id: path.id
          model_memory_limit: requestBody.model_memory_limit
      - text: Change the description of analytics job {id} to {description}.
        slots:
          id: path.id
          description: requestBody.description
      method: generated
      generated: '2026-09-26'