sources:
  - name: s3
    config:
      group: rds
    tables:
      - name: bots_generatedclip_insert
        description: |
          Raw bots_generatedclip data from S3 containing clip generation data exported from RDS.
          This table stores parquet files exported from the RDS bots_generatedclip table to S3 for analytics and reporting purposes.
          Used to enable Dagster dependency on the Glue job that exports this data and provides access to
          raw clip generation data including creation details, metadata, and generation parameters.
        tags: ['verified', 'partition_type=hourly']
        columns:
          - name: id
            description: "Clip ID (primary key)"
          - name: created_at
            description: "Timestamp when the clip was created"
          - name: user_id
            description: "User ID who created the clip"
          - name: request_id
            description: "Request ID for the generation"
          - name: s3_id
            description: "S3 object identifier for the clip's audio file"
          - name: status
            description: "Status of the clip (e.g., complete, failed)"
          - name: batch_index
            description: "Index within batch generation request"
          - name: title
            description: "Title of the clip"
          - name: time_used
            description: "Time in seconds taken to generate"
          - name: prompt_text
            description: "Prompt text used for generation"
          - name: model_name
            description: "Model name used for generation"
          - name: metadata
            description: "JSON metadata containing additional generation parameters"
          - name: p_date
            description: "Partition date parsed from file path"
          - name: p_hour
            description: "Partition hour parsed from file path"
        meta:
          dagster:
            asset_key: ['clip_insert']