Skip to content

Input Parameters

Input parameters define the data that your tasks will process. Understanding how to configure them correctly is crucial for building effective workflows.

Parameter Types

Type Description Example
STRING Text data "Hello World"
INTEGER Whole numbers 42
FLOAT Decimal numbers 3.14
BOOLEAN True/false values true
Type Description Example Restrictions
ARRAY List of items ["item1", "item2"] Maximum 2D arrays allowed
OBJECT Structured data {"name": "John"} Properties must be basic types
ENUM Limited choices "APPROVED" Fixed set of values
Type Description Use Case
OUTPUT_ARTIFACT Reference to task output Inter-task data flow
KNOWLEDGE_ARTIFACT Knowledge base reference Accessing stored knowledge

Complex Type Configuration

Array Parameters

⚠️ Important Rules:

  • Maximum 2D arrays allowed
  • Inner array elements can be basic types or objects
  • No 3D or deeper arrays

Object Parameters

⚠️ Important Rules:

  • Properties must use only:
  • STRING
  • INTEGER
  • FLOAT
  • BOOLEAN
  • No nested objects allowed
  • Each property needs name, description, type, and required fields

Array Structure

// 1D Array
{
  "type": "array",
  "items": {
    "type": "string",  // Any basic type
    "items": null,
    "properties": null
  },
  "properties": null
}

// 2D Array (Maximum depth)
{
  "type": "array",
  "items": {
    "type": "array",
    "items": {
      "type": "string",  // Any basic type
      "items": null,
      "properties": null
    },
    "properties": null
  },
  "properties": null
}

// Array of Objects
{
  "type": "array",
  "items": {
    "type": "object",
    "items": null,
    "properties": [
      {
        "name": "field",
        "description": "Description",
        "required": false,
        "type": "string"  // Must be basic type
      }
    ]
  },
  "properties": null
}

Object Structure

{
  "type": "object",
  "items": null,
  "properties": [
    {
      "name": "fieldName",
      "description": "Field description",
      "required": false,
      "type": "string"  // Must be one of: "string", "integer", "float", "boolean"
    }
  ]
}

Parameter Sources

Available Sources

  1. Task Config

    • Set during task setup
    • Fixed values
    • Template values

  2. System Generated

    • Created during execution
    • Dynamic values
    • No manual input needed

  3. Human Input

    • Provided during workflow
    • User interaction required
    • Interactive forms
// Task Config
{
  "source": "task_config",
  "value": "configured_value"
}

// System Generated
{
  "source": "system_generated"
}

// Human Input
{
  "source": "human_input"
}

Data Sources

Memory Storage

  • Default storage method
  • For regular data volumes
  • No special configuration needed

Data Lake Storage

  • For large datasets
  • Persistent storage
  • Requires data_lake_id
// Memory Storage
{
  "data_source": "MEMORY"
}

// Data Lake Storage
{
  "data_source": "DATA_LAKE",
  "data_lake_id": "12345"
}

Configuration Examples

String Parameter

{
  "name": "username",
  "description": "User's login name",
  "type": "STRING",
  "required": true,
  "source": "task_config",
  "value": "john_doe"
}

Number Parameter

{
  "name": "age",
  "description": "User's age",
  "type": "INTEGER",
  "required": true,
  "source": "task_config",
  "value": 25
}

2D Array Example

{
  "name": "matrix",
  "type": "array",
  "items": {
    "type": "array",
    "items": {
      "type": "integer",
      "items": null,
      "properties": null
    },
    "properties": null
  },
  "properties": null,
  "description": "2D matrix of values",
  "required": true
}

Object Example

{
  "name": "userProfile",
  "type": "object",
  "items": null,
  "properties": [
    {
      "name": "firstName",
      "type": "string",
      "description": "User's first name",
      "required": true
    },
    {
      "name": "age",
      "type": "integer",
      "description": "User's age",
      "required": false
    }
  ]
}

Validation Rules

  1. Object Properties

    • Only basic types allowed (string, integer, float, boolean)
    • All fields required (name, description, type, required)
    • No nested objects

  2. Array Items

    • Maximum 2D arrays
    • Basic types or objects for elements
    • Proper null values for unused fields

  3. Type Consistency

    • Must follow exact schema
    • No additional fields
    • Proper null handling
// ❌ 3D Array (Not Allowed)
{
  "type": "array",
  "items": {
    "type": "array",
    "items": {
      "type": "array",  // Third level not allowed
      "items": {
        "type": "string"
      }
    }
  }
}

// ❌ Nested Object (Not Allowed)
{
  "type": "object",
  "properties": [
    {
      "name": "address",
      "type": "object",  // Cannot use object type in properties
      "required": true
    }
  ]
}

Best Practices

  1. Naming Conventions

    • Use clear, descriptive names
    • Follow consistent conventions
    • Indicate purpose in name

  2. Documentation

    • Be specific and clear
    • Include format requirements
    • Note any dependencies

  3. Type Selection

    • Use simplest type possible
    • Consider data flow requirements
    • Plan for scale

  4. Validation

    • Set appropriate required flags
    • Include format constraints
    • Consider edge cases

Common Issues and Solutions

Issue Solution
Type mismatch Verify data type matches configuration
Missing required value Check source configuration
Invalid format Review validation rules
Data source error Verify data lake configuration
Invalid object property Use only allowed basic types
Array depth exceeded Restructure to 2D maximum