Oxen Data Frames API

openapi: 3.0.0
info:
  title: oxen Ai Data Frames API
  version: 0.243.1
servers:
- url: https://hub.oxen.ai
  variables: {}
security: []
tags:
- name: Data Frames
paths:
  /api/repos/{namespace}/{repo_name}/data_frames/{resource}:
    get:
      description: Get a paginated slice of a tabular data frame with optional filtering and transformations.
      operationId: get
      parameters:
      - description: Namespace of the repository
        example: ox
        in: path
        name: namespace
        required: true
        schema:
          type: string
      - description: Name of the repository
        example: ImageNet-1k
        in: path
        name: repo_name
        required: true
        schema:
          type: string
      - description: Path to the tabular file (including branch/commit info)
        example: main/data/labels.csv
        in: path
        name: resource
        required: true
        schema:
          type: string
      - in: path
        name: columns
        required: true
        schema:
          type:
          - string
          - 'null'
      - in: path
        name: delimiter
        required: true
        schema:
          type:
          - string
          - 'null'
      - in: path
        name: find_embedding_where
        required: true
        schema:
          type:
          - string
          - 'null'
      - in: path
        name: filter
        required: true
        schema:
          type:
          - string
          - 'null'
      - in: path
        name: output
        required: true
        schema:
          type:
          - string
          - 'null'
      - in: path
        name: output_column
        required: true
        schema:
          type:
          - string
          - 'null'
      - in: path
        name: page_size
        required: true
        schema:
          minimum: 0
          type:
          - integer
          - 'null'
      - in: path
        name: page
        required: true
        schema:
          minimum: 0
          type:
          - integer
          - 'null'
      - in: path
        name: row
        required: true
        schema:
          minimum: 0
          type:
          - integer
          - 'null'
      - in: path
        name: randomize
        required: true
        schema:
          type:
          - boolean
          - 'null'
      - in: path
        name: reverse
        required: true
        schema:
          type:
          - boolean
          - 'null'
      - in: path
        name: slice
        required: true
        schema:
          type:
          - string
          - 'null'
      - in: path
        name: sort_by
        required: true
        schema:
          type:
          - string
          - 'null'
      - in: path
        name: sort_by_similarity_to
        required: true
        schema:
          type:
          - string
          - 'null'
      - in: path
        name: sql
        required: true
        schema:
          type:
          - string
          - 'null'
      - in: path
        name: take
        required: true
        schema:
          type:
          - string
          - 'null'
      responses:
        '200':
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/JsonDataFrameViewResponse'
          description: Data frame slice found
        '404':
          description: File or resource not found
      summary: Get data frame slice
      tags:
      - Data Frames
  /api/repos/{namespace}/{repo_name}/compare/data_frames:
    post:
      description: Create and cache a tabular diff comparing two data frames with configurable keys and target columns.
      operationId: create_df_diff
      parameters:
      - description: Namespace of the repository
        example: ox
        in: path
        name: namespace
        required: true
        schema:
          type: string
      - description: Name of the repository
        example: compare-datasets
        in: path
        name: repo_name
        required: true
        schema:
          type: string
      requestBody:
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/TabularCompareBody'
        description: Tabular comparison configuration (files, revisions, keys, targets)
        required: true
      responses:
        '200':
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/CompareTabularResponse'
          description: Data frame diff created and cached successfully
        '400':
          description: Bad Request (e.g., failed to parse body)
        '404':
          description: Repository, revision, or file not found
      summary: Create a tabular data frame diff
      tags:
      - Data Frames
  /api/repos/{namespace}/{repo_name}/compare/data_frames/{compare_id}:
    delete:
      description: Delete a cached tabular diff comparison by its ID.
      operationId: delete_df_diff
      parameters:
      - description: Namespace of the repository
        example: ox
        in: path
        name: namespace
        required: true
        schema:
          type: string
      - description: Name of the repository
        example: compare-datasets
        in: path
        name: repo_name
        required: true
        schema:
          type: string
      - description: The unique ID of the comparison session to delete
        example: uuid-1234
        in: path
        name: compare_id
        required: true
        schema:
          type: string
      responses:
        '200':
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/StatusMessage'
          description: Data frame diff deleted successfully
        '404':
          description: Repository or comparison ID not found
      summary: Delete DF Diff
      tags:
      - Data Frames
    get:
      description: Retrieve a previously cached tabular diff by its comparison ID.
      operationId: get_df_diff
      parameters:
      - description: Namespace of the repository
        example: ox
        in: path
        name: namespace
        required: true
        schema:
          type: string
      - description: Name of the repository
        example: compare-datasets
        in: path
        name: repo_name
        required: true
        schema:
          type: string
      - description: The unique ID of the comparison session
        example: uuid-1234
        in: path
        name: compare_id
        required: true
        schema:
          type: string
      - description: The base and head revisions separated by '..'
        example: main..feature/add-labels
        in: path
        name: base_head
        required: true
        schema:
          type: string
      requestBody:
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/TabularCompareBody'
        description: Tabular comparison information needed to fetch the correct cached diff (files, revisions)
        required: true
      responses:
        '200':
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/CompareTabularResponse'
          description: Cached data frame diff found successfully
        '404':
          description: Cached diff not found for the given ID and entries
      summary: Get a cached tabular data frame diff
      tags:
      - Data Frames
    put:
      description: Update an existing cached tabular diff comparison with new configuration.
      operationId: update_df_diff
      parameters:
      - description: Namespace of the repository
        example: ox
        in: path
        name: namespace
        required: true
        schema:
          type: string
      - description: Name of the repository
        example: compare-datasets
        in: path
        name: repo_name
        required: true
        schema:
          type: string
      - description: The unique ID of the comparison session
        example: uuid-1234
        in: path
        name: compare_id
        required: true
        schema:
          type: string
      requestBody:
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/TabularCompareBody'
        description: Updated tabular comparison configuration
        required: true
      responses:
        '200':
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/CompareTabularResponse'
          description: Data frame diff updated and cached successfully
        '400':
          description: Bad Request
        '404':
          description: Comparison ID, Repository, revision, or file not found
      summary: Update tabular data frame diff
      tags:
      - Data Frames
  /api/repos/{namespace}/{repo_name}/data_frames/from_directory/{resource}:
    post:
      description: Create a data frame by scanning directory contents and commit it to a branch.
      operationId: from_directory
      parameters:
      - description: Namespace of the repository
        example: ox
        in: path
        name: namespace
        required: true
        schema:
          type: string
      - description: Name of the repository
        example: CattleData
        in: path
        name: repo_name
        required: true
        schema:
          type: string
      - description: Directory path to read from (including branch/commit info)
        example: main/data/images
        in: path
        name: resource
        required: true
        schema:
          type: string
      requestBody:
        content:
          application/json:
            example:
              commit_message: Generated image index
              extra_columns:
              - size
              - extension
              output_path: data/image_index.csv
              recursive: true
              user_email: bessie@oxen.ai
              user_name: Bessie
            schema:
              $ref: '#/components/schemas/FromDirectoryRequest'
        description: Options for creating a data frame from a directory, including output path, columns, and commit message.
        required: true
      responses:
        '200':
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/CommitResponse'
          description: Data frame created and committed
        '400':
          description: Invalid request body or resource path
        '404':
          description: Repository or resource not found
      summary: Create data frame from directory
      tags:
      - Data Frames
  /api/repos/{namespace}/{repo_name}/data_frames/{resource}/index:
    post:
      description: Start indexing a tabular file for queryable access. Creates a workspace if the file is not already indexed.
      operationId: index
      parameters:
      - description: Namespace of the repository
        example: ox
        in: path
        name: namespace
        required: true
        schema:
          type: string
      - description: Name of the repository
        example: CattleData
        in: path
        name: repo_name
        required: true
        schema:
          type: string
      - description: Path to the tabular file to index (including branch/commit info)
        example: main/data/weights.csv
        in: path
        name: resource
        required: true
        schema:
          type: string
      responses:
        '200':
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/StatusMessage'
          description: Indexing process started or completed
        '404':
          description: Resource not found
        '409':
          description: Dataset already indexed
      summary: Start data frame indexing
      tags:
      - Data Frames
components:
  schemas:
    NewColumn:
      properties:
        data_type:
          type: string
        name:
          type: string
      required:
      - name
      - data_type
      type: object
    TabularCompareBody:
      properties:
        compare:
          items:
            $ref: '#/components/schemas/TabularCompareTargetBody'
          type: array
        compare_id:
          type: string
        display:
          items:
            $ref: '#/components/schemas/TabularCompareTargetBody'
          type: array
        keys:
          items:
            $ref: '#/components/schemas/TabularCompareFieldBody'
          type: array
        left:
          $ref: '#/components/schemas/TabularCompareResourceBody'
        right:
          $ref: '#/components/schemas/TabularCompareResourceBody'
      required:
      - compare_id
      - left
      - right
      - keys
      - compare
      - display
      type: object
    CompareDupes:
      properties:
        left:
          format: int64
          minimum: 0
          type: integer
        right:
          format: int64
          minimum: 0
          type: integer
      required:
      - left
      - right
      type: object
    CompareSchemaColumn:
      properties:
        dtype:
          type: string
        key:
          type: string
        name:
          type: string
      required:
      - name
      - key
      - dtype
      type: object
    MessageLevel:
      enum:
      - Info
      - Warning
      - Error
      type: string
    DFOptsView:
      properties:
        opts:
          items:
            $ref: '#/components/schemas/DFOptView'
          type: array
      required:
      - opts
      type: object
    JsonDataFrameViews:
      properties:
        source:
          $ref: '#/components/schemas/DataFrameSchemaSize'
        view:
          $ref: '#/components/schemas/JsonDataFrameView'
      required:
      - source
      - view
      type: object
    DerivedDFResource:
      properties:
        path:
          type: string
        resource_id:
          type: string
        resource_type:
          $ref: '#/components/schemas/DFResourceType'
      required:
      - resource_id
      - path
      - resource_type
      type: object
    CompareTabularResponse:
      allOf:
      - $ref: '#/components/schemas/StatusMessage'
      - properties:
          dfs:
            $ref: '#/components/schemas/CompareTabular'
          messages:
            items:
              $ref: '#/components/schemas/OxenMessage'
            type: array
        required:
        - dfs
        - messages
        type: object
    CompareTabular:
      properties:
        display:
          items:
            $ref: '#/components/schemas/TabularCompareTargetBody'
          type:
          - array
          - 'null'
        dupes:
          $ref: '#/components/schemas/CompareDupes'
        keys:
          items:
            $ref: '#/components/schemas/TabularCompareFieldBody'
          type:
          - array
          - 'null'
        schema_diff:
          oneOf:
          - type: 'null'
          - $ref: '#/components/schemas/CompareSchemaDiff'
        source_schemas:
          $ref: '#/components/schemas/CompareSourceSchemas'
        summary:
          oneOf:
          - type: 'null'
          - $ref: '#/components/schemas/CompareSummary'
        targets:
          items:
            $ref: '#/components/schemas/TabularCompareTargetBody'
          type:
          - array
          - 'null'
      required:
      - dupes
      - source_schemas
      type: object
    CommitResponse:
      allOf:
      - $ref: '#/components/schemas/StatusMessage'
      - properties:
          commit:
            $ref: '#/components/schemas/Commit'
        required:
        - commit
        type: object
      example:
        commit:
          author: ox
          email: ox@example.com
          id: a1b2c3d4e5f67890abcdef1234567890
          message: Refactor data loading pipeline.
          parent_ids:
          - f1e2d3c4b5a67890fedcba9876543210
          timestamp: '2025-01-01T10:00:00Z'
        status: success
        status_message: resource_found
    CompareSchemaDiff:
      properties:
        added_cols:
          items:
            $ref: '#/components/schemas/CompareSchemaColumn'
          type: array
        removed_cols:
          items:
            $ref: '#/components/schemas/CompareSchemaColumn'
          type: array
      required:
      - added_cols
      - removed_cols
      type: object
    TabularCompareFieldBody:
      properties:
        alias_as:
          type:
          - string
          - 'null'
        compare_method:
          type:
          - string
          - 'null'
        left:
          type: string
        right:
          type: string
      required:
      - left
      - right
      type: object
    Pagination:
      properties:
        page_number:
          minimum: 0
          type: integer
        page_size:
          minimum: 0
          type: integer
        total_entries:
          minimum: 0
          type: integer
        total_pages:
          minimum: 0
          type: integer
      required:
      - page_size
      - page_number
      - total_pages
      - total_entries
      type: object
    DataFrameSchemaSize:
      properties:
        schema:
          $ref: '#/components/schemas/Schema'
        size:
          $ref: '#/components/schemas/DataFrameSize'
      required:
      - schema
      - size
      type: object
    Changes:
      properties:
        previous:
          oneOf:
          - type: 'null'
          - $ref: '#/components/schemas/PreviousField'
        status:
          type: string
      required:
      - status
      type: object
    Schema:
      properties:
        fields:
          items:
            $ref: '#/components/schemas/Field'
          type: array
        hash:
          type: string
        metadata: {}
      required:
      - hash
      - fields
      type: object
    DFResourceType:
      enum:
      - compare
      - diff
      - query
      type: string
    DataFrameSize:
      properties:
        height:
          minimum: 0
          type: integer
        width:
          minimum: 0
          type: integer
      required:
      - height
      - width
      type: object
    Field:
      properties:
        changes:
          oneOf:
          - type: 'null'
          - $ref: '#/components/schemas/Changes'
        dtype:
          type: string
        metadata: {}
        name:
          type: string
      required:
      - name
      - dtype
      type: object
    CompareTabularMods:
      properties:
        added_rows:
          minimum: 0
          type: integer
        modified_rows:
          minimum: 0
          type: integer
        removed_rows:
          minimum: 0
          type: integer
      required:
      - added_rows
      - removed_rows
      - modified_rows
      type: object
    TabularCompareResourceBody:
      properties:
        path:
          type: string
        version:
          type: string
      required:
      - path
      - version
      type: object
    StatusMessage:
      properties:
        oxen_version:
          type:
          - string
          - 'null'
        status:
          type: string
        status_message:
          type: string
      required:
      - status
      - status_message
      type: object
    PreviousField:
      properties:
        dtype:
          type: string
        metadata: {}
        name:
          type: string
      required:
      - name
      - dtype
      type: object
    FromDirectoryRequest:
      properties:
        commit_message:
          type:
          - string
          - 'null'
        extra_columns:
          items:
            $ref: '#/components/schemas/NewColumn'
          type:
          - array
          - 'null'
        output_path:
          type:
          - string
          - 'null'
        recursive:
          type:
          - boolean
          - 'null'
        user_email:
          type:
          - string
          - 'null'
        user_name:
          type:
          - string
          - 'null'
      type: object
    JsonDataFrameView:
      allOf:
      - $ref: '#/components/schemas/DFOptsView'
      - properties:
          data: {}
          pagination:
            $ref: '#/components/schemas/Pagination'
          schema:
            $ref: '#/components/schemas/Schema'
          size:
            $ref: '#/components/schemas/DataFrameSize'
        required:
        - schema
        - size
        - data
        - pagination
        type: object
    Commit:
      example:
        author: ox
        email: ox@example.com
        id: a1b2c3d4e5f67890abcdef1234567890
        message: Refactor data loading pipeline.
        parent_ids:
        - f1e2d3c4b5a67890fedcba9876543210
        timestamp: '2025-01-01T10:00:00Z'
      properties:
        author:
          type: string
        email:
          type: string
        id:
          type: string
        message:
          type: string
        parent_ids:
          items:
            type: string
          type: array
        timestamp:
          format: date-time
          type: string
      required:
      - id
      - parent_ids
      - message
      - author
      - email
      - timestamp
      type: object
    JsonDataFrameViewResponse:
      allOf:
      - $ref: '#/components/schemas/StatusMessage'
      - properties:
          commit:
            oneOf:
            - type: 'null'
            - $ref: '#/components/schemas/Commit'
          data_frame:
            $ref: '#/components/schemas/JsonDataFrameViews'
          derived_resource:
            oneOf:
            - type: 'null'
            - $ref: '#/components/schemas/DerivedDFResource'
          resource:
            oneOf:
            - type: 'null'
            - $ref: '#/components/schemas/ResourceVersion'
        required:
        - data_frame
        type: object
    DFOptView:
      properties:
        name:
          type: string
        value: {}
      required:
      - name
      - value
      type: object
    ResourceVersion:
      properties:
        path:
          type: string
        version:
          type: string
      required:
      - path
      - version
      type: object
    CompareSourceSchemas:
      properties:
        left:
          $ref: '#/components/schemas/Schema'
        right:
          $ref: '#/components/schemas/Schema'
      required:
      - left
      - right
      type: object
    TabularCompareTargetBody:
      properties:
        compare_method:
          type:
          - string
          - 'null'
        left:
          type:
          - string
          - 'null'
        right:
          type:
          - string
          - 'null'
      type: object
    CompareSummary:
      properties:
        modifications:
          $ref: '#/components/schemas/CompareTabularMods'
        schema:
          $ref: '#/components/schemas/Schema'
      required:
      - modifications
      - schema
      type: object
    OxenMessage:
      properties:
        description:
          type: string
        level:
          $ref: '#/components/schemas/MessageLevel'
        title:
          type: string
      required:
      - level
      - title
      - description
      type: object
  securitySchemes:
    authorization:
      scheme: bearer
      type: http