validate_schema

Validate context data against a JSON Schema.

Parameters

*Provide exactly one of schema or schema_key.

The node retrieves the value stored under source_key in the workflow context and validates it using the provided JSON Schema. If validation fails, the node returns an error and the workflow step is marked as failed.

Use this node when the value is already in your context as JSON (object/array/value).
If the context value is a raw JSON string, use json_validate instead.

Context Output

• validation_success -- boolean indicating whether validation passed
• validation_errors -- array of error description strings (empty on success)

Note: When validation fails the node returns a structured error. Downstream steps execute only if a dedicated on_error handler successfully resolves that failure. After retries are exhausted, the final validation result is buffered for the source phase barrier and the handler receives its source-exact copy in _error_output. Normal shared keys still follow declaration-order collision precedence if a parallel step publishes the same names.

Example

local flow = Flow.new("user_registration")

flow:step("validate", nodes.validate_schema({
    source_key = "payload",
    schema = {
        type = "object",
        required = { "email", "name" },
        properties = {
            email = { type = "string" },
            name  = { type = "string" },
            age   = { type = "integer", minimum = 0 }
        }
    }
}))

flow:step("done", nodes.log({
    message = "Validation passed for ${ctx.payload.name} (${ctx.payload.email})"
})):depends_on("validate")

return flow

Validate with a schema loaded from context

local flow = Flow.new("validate_from_context")

flow:step("read_schema", nodes.read_file({
    path = "schemas/user.schema.json",
    output_key = "schema"
}))

flow:step("validate", nodes.validate_schema({
    source_key = "payload",
    schema_key = "schema_content"
})):depends_on("read_schema")

return flow