Validate Schema

Type key: validate_schema_node
Group: Data
Category: data
Tier: CORE

The Validate Schema node checks incoming data against a set of validation rules to ensure data quality before processing. It can halt the workflow on invalid data or continue with warnings, letting downstream nodes decide how to handle errors.

Handles

Handle	Direction	Description
`input`	Input	Receives the data to validate
`output`	Output	Passes validated data (behavior depends on mode)

Configuration

Field	Type	Default	Description
`mode`	enum	`fail`	Behavior on validation failure
`rules`	array	`[]`	Visual validation rules (no-code)
`schema`	string	`""`	Full JSON Schema for advanced validation

Validation Mode

Mode	Behavior on Invalid Data
`fail`	Halt the workflow immediately with a `VALIDATION_ERROR`
`soft_fail`	Continue execution; attach validation errors to output; let downstream decide

Visual Rule Builder

The rule builder provides a no-code interface for defining validation constraints. Each rule validates one field from the incoming data.

Rule Object

{
  "field": "email",
  "type": "email",
  "required": true,
  "min": "",
  "max": "",
  "pattern": ""
}

Property	Description
`field`	Field name to validate (supports dot notation for nested)
`type`	Expected data type — see table below
`required`	If `true`, the field must be present and non-empty
`min`	Minimum value (number/integer) or minimum length (string)
`max`	Maximum value (number/integer) or maximum length (string)
`pattern`	Regex pattern — only for `string` type (optional)

Supported Types

Type	Validates	Min/Max meaning	Pattern
`string`	Must be a string	Character length	yes
`number`	Must be numeric	Value range	no
`integer`	Must be an integer	Value range	no
`boolean`	Must be boolean	—	no
`email`	Must be a valid email address	—	no
`url`	Must be a valid URL	—	no
`date`	Must be a parseable date string	—	no
`array`	Must be an array	—	no

Upstream Field Discovery

The Field dropdown automatically discovers available fields from upstream nodes. When connected to a Dummy Data or Data Model node, all fields are listed for quick selection. If the upstream data is not yet available (e.g. after importing a JSON preset), the previously saved field names are preserved and shown in the dropdown.

UI Layout

Each rule is displayed as a card with hover-to-delete (X button, top-right corner):

Field — full-width select with upstream field discovery
Type + Required checkbox — side by side
Min / Max — side by side (shown only for string, number, integer)
Regex pattern — shown only for string type

JSON Schema (Advanced)

For complex validations beyond the visual rule builder, use a full JSON Schema (Draft-7). When a schema is configured and no visual rules exist, the node falls back to legacy JSON Schema validation.

Example: Validate an Object

{
  "type": "object",
  "required": ["email", "age"],
  "properties": {
    "email": { "type": "string", "format": "email" },
    "age": { "type": "integer", "minimum": 18, "maximum": 120 },
    "name": { "type": "string", "minLength": 2, "maxLength": 100 }
  },
  "additionalProperties": false
}

Example: Validate an Array of Products

{
  "type": "array",
  "minItems": 1,
  "items": {
    "type": "object",
    "required": ["sku", "price"],
    "properties": {
      "sku": { "type": "string", "minLength": 3 },
      "price": { "type": "number", "exclusiveMinimum": 0 },
      "stock": { "type": "integer", "minimum": 0 }
    }
  }
}

Output

On Success (`mode: fail`)

Input passes through unchanged. A nodes.{nodeId} metadata entry is added confirming validation passed.

On Success (`mode: soft_fail`)

{
  "...original fields...",
  "nodes": {
    "validate_schema_node-001": {
      "type": "validation",
      "validated": true,
      "valid": true,
      "rules_count": 3,
      "mode": "soft_fail"
    }
  }
}

On Failure (`mode: soft_fail`)

{
  "...original fields...",
  "nodes": {
    "validate_schema_node-001": {
      "type": "validation",
      "validated": true,
      "valid": false,
      "errors": [
        "Field 'email' is required",
        "Field 'age' must be at least 18"
      ],
      "warnings": ["Field 'email' is required", "Field 'age' must be at least 18"],
      "rules_count": 3,
      "mode": "soft_fail"
    }
  }
}

On Failure (`mode: fail`)

The workflow halts with a VALIDATION_ERROR. No output is produced.

Example Workflow: Webhook Input Validation

A Dummy Data node simulates an incoming webhook payload. The Validate Schema node checks that name is a string of 2–100 characters, email is a valid email address, and amount is a number ≥ 0.01. A Data Viewer shows the validated result.

Validate Schema node — webhook input validation

::: info 💾 Try this example Download the workflow JSON — open an empty workflow and import the file. :::

Use Cases

Scenario	Rules
Webhook input validation	`name` string required, `email` email required
Order processing	`amount` number ≥ 0.01, `currency` string required
User registration	`age` integer ≥ 18, `email` email required
API payload sanitization	`sku` string pattern `^[A-Z]{3}-\d+$`
Form submission	`phone` string, `message` string min 10 chars

Notes

Use mode: fail when invalid data should never reach downstream nodes (defensive pipelines)
Use mode: soft_fail when you want to log or quarantine invalid data while allowing the workflow to continue
The field selector supports dot notation for nested fields (e.g. address.city, payment.method)
When validating an array of items, each item is validated individually and errors include the item index
Visual rules take priority over legacy JSON Schema — if both are configured, rules are used