Skip to main content
Field mappings and transformations let you reshape data between your workflow and external services without writing code. They are defined inside the data object of executor nodes in your workflow definition.
These fields live inside the node’s data property — a flexible key-value object that varies by node type. They are not separate top-level fields in the workflow schema.

​
Input mapping

Input mapping converts workflow fields into the fields expected by an executor. Flowker applies input mappings before calling the external service. Define an inputMapping array in the executor node’s data object. Each entry specifies a source (a workflow field path) and a target (the executor field it maps to). 
{
  "id": "executor-balance",
  "type": "executor",
  "name": "Check Balance",
  "data": {
    "providerConfigId": "a1b2c3d4-e5f6-4789-a012-345678901234",
    "inputMapping": [
      { "source": "workflow.customerId", "target": "executor.accountId" },
      { "source": "workflow.amount", "target": "executor.minimumBalance" }
    ]
  }
}

​
Mapping entry fields

FieldTypeRequiredDescription
sourcestringYesJSONPath to the source value (e.g., workflow.customer.cpf).
targetstringYesJSONPath to the target field (e.g., executor.document).
transformationobjectNoOptional transformation to apply during mapping. See Inline transformations.
requiredbooleanNoIf true, the mapping fails when the source path doesn’t exist. Defaults to false.

​
Output mapping

Output mapping extracts fields from the executor’s response and writes them back into the workflow context. Downstream nodes can then access the mapped values. Flowker applies output mappings after receiving the response. 
{
  "data": {
    "outputMapping": [
      { "source": "executor.currentBalance", "target": "workflow.balance" },
      { "source": "executor.accountStatus", "target": "workflow.status" }
    ]
  }
}
The same entry fields apply as input mapping (source, target, transformation, required).

​
Inline transformations

When a simple field-to-field mapping isn’t enough, attach a transformation object to a mapping entry. The transformation is applied to the source value before writing it to the target. 
{
  "source": "workflow.phone",
  "target": "executor.phoneNumber",
  "transformation": {
    "type": "remove_characters",
    "config": {
      "characters": ".- ()"
    }
  }
}

​
Available transformation types

TypeWhat it doesConfig fields
remove_charactersStrips specified characters from a string.characters — string of characters to remove.
add_prefixPrepends a string to a field value.prefix — string to prepend.
add_suffixAppends a string to a field value.suffix — string to append.
to_uppercaseConverts a string to uppercase.—
to_lowercaseConverts a string to lowercase.—

​
Kazaam transformations

For advanced JSON-to-JSON transformations that go beyond field mapping, define a transforms array in the executor node’s data object. These use the Kazaam transformation engine. 
{
  "data": {
    "transforms": [
      {
        "operation": "shift",
        "spec": {
          "outputField": "inputField"
        }
      }
    ]
  }
}

​
Kazaam operation fields

FieldTypeRequiredDescription
operationstringYesThe Kazaam operation type (e.g., shift, concat, remove_characters).
specobjectYesOperation-specific configuration.
requirebooleanNoIf true, all paths referenced in spec must exist. Defaults to false.
Kazaam supports operations like shift (move/rename fields), concat (combine fields), coalesce (first non-null value), and default (set fallback values). The custom operations listed in inline transformations are also available as Kazaam operations.
When using custom transforms like remove_characters as Kazaam operations, the spec object must include a path field pointing to the JSON path of the data to transform. This differs from inline transformations, where the path is derived automatically from the source/target mapping.

​
Execution order

When Flowker executes an executor node:
  1. Input mappings are applied first — workflow data is mapped to executor input format.
  2. Kazaam transforms run next on the mapped input (if defined).
  3. The executor call is made to the external service.
  4. Output mappings are applied to the response — executor data is mapped back to the workflow context.

​
Troubleshooting

If a transformation fails during execution, the step is marked as failed with a message indicating which mapping caused the error. Common issues:
  • Source path doesn’t exist — check that the upstream node actually produces the field you’re referencing. Set required: true to catch missing fields early.
  • Type mismatch — string-to-number conversions are not automatic. Use a transformation to convert formats.
  • Invalid Kazaam spec — verify the operation name and spec structure against the Kazaam documentation.