tool_adapters/datapipe.yaml

name: datapipe
description: Data pipeline orchestrator for connecting sources, applying transformations, and routing to sinks on configurable schedules
binary: datapipe
auth:
  type: env_var
  key: DATAPIPE_API_KEY
commands:
  - name: source list
    description: List all configured data sources
    args:
      - name: type
        type: enum
        required: false
        description: Filter by source type
        values: ["postgres", "mysql", "s3", "api", "kafka", "mongodb"]
    output_format: json
    side_effects: false
    example: "datapipe source list --type postgres"

  - name: source connect
    description: Register a new data source connection
    args:
      - name: name
        type: string
        required: true
        description: Source name (unique identifier)
      - name: type
        type: enum
        required: true
        description: Source type
        values: ["postgres", "mysql", "s3", "api", "kafka", "mongodb"]
      - name: config
        type: json
        required: true
        description: "Connection configuration as JSON (e.g. {\"host\": \"db.example.com\", \"port\": 5432, \"database\": \"analytics\"})"
    output_format: json
    side_effects: true
    example: "datapipe source connect --name prod-db --type postgres --config '{\"host\": \"db.example.com\", \"port\": 5432, \"database\": \"analytics\"}'"

  - name: transform create
    description: Create a named transformation step with SQL logic
    args:
      - name: name
        type: string
        required: true
        description: Transform name (unique identifier)
      - name: sql
        type: string
        required: true
        description: SQL transformation query
      - name: source
        type: string
        required: true
        description: Source name to read data from
      - name: description
        type: string
        required: false
        description: Human-readable description of the transformation
    output_format: json
    side_effects: true
    example: "datapipe transform create --name daily-revenue --sql 'SELECT date, SUM(amount) as revenue FROM orders GROUP BY date' --source prod-db"

  - name: transform run
    description: Execute a transformation step
    args:
      - name: name
        type: string
        required: true
        description: Transform name to execute
      - name: dry-run
        type: bool
        required: false
        description: Preview the output without writing results
      - name: limit
        type: int
        required: false
        description: Limit output rows (useful for preview)
    output_format: json
    side_effects: true
    example: "datapipe transform run --name daily-revenue --dry-run --limit 10"

  - name: sink create
    description: Create a data sink destination
    args:
      - name: name
        type: string
        required: true
        description: Sink name (unique identifier)
      - name: type
        type: enum
        required: true
        description: Sink type
        values: ["s3", "bigquery", "warehouse", "postgres", "elasticsearch"]
      - name: config
        type: json
        required: true
        description: "Sink configuration as JSON (e.g. {\"bucket\": \"analytics-output\", \"prefix\": \"daily/\"})"
    output_format: json
    side_effects: true
    example: "datapipe sink create --name analytics-lake --type s3 --config '{\"bucket\": \"analytics-output\", \"prefix\": \"daily/\"}'"

  - name: pipeline create
    description: Create a complete data pipeline connecting source, transforms, and sink with a schedule
    args:
      - name: name
        type: string
        required: true
        description: Pipeline name (unique identifier)
      - name: source
        type: string
        required: true
        description: Source name
      - name: transforms
        type: string
        required: true
        description: Comma-separated list of transform names in execution order
      - name: sink
        type: string
        required: true
        description: Sink name
      - name: schedule
        type: string
        required: true
        description: Cron expression for scheduled execution
    output_format: json
    side_effects: true
    example: "datapipe pipeline create --name daily-etl --source prod-db --transforms daily-revenue,add-dimensions --sink analytics-lake --schedule '0 2 * * *'"

  - name: pipeline status
    description: Check the status and execution history of a pipeline
    args:
      - name: name
        type: string
        required: true
        description: Pipeline name
    output_format: json
    side_effects: false
    example: "datapipe pipeline status --name daily-etl"