Target Fields Overview
Target fields define the structured output schema for the FastGPT content extraction workflow node. Each configured target field generates a unique named output variable, enabling downstream workflow nodes to reference extracted data directly. This configuration converts unstructured input text into a standardized, machine-readable format for consistent processing.
Target Field Parameter Reference
The following table lists all official parameters for target field configuration, as specified in FastGPT documentation:
| Parameter Name | Official Purpose | Critical Constraints |
|---|---|---|
| Key | Unique identifier for the extraction field | Must not be duplicated across all target fields in a single content extract node |
| Field Description | Describes the specific data the field is intended to capture | Examples include product name, event timestamp, search keyword, or similar identifiers |
| Required | Indicates if the language model is obligated to attempt extraction for the field | Even when enabled, the model may return an empty string if no matching data exists in the input |
Step-by-Step Configuration
Follow these official steps to set up target fields for a content extraction node:
- Open the FastGPT workflow editor and add a content extraction node to your workflow.
- Access the node’s configuration panel to locate the Target Fields configuration section.
- Click the add button to create a new target field entry.
- Assign a unique, non-duplicate key to the new field to avoid configuration conflicts.
- Write a clear field description to specify the type of data the field should extract.
- Toggle the Required switch to set whether the model must attempt extraction for this field.
- Repeat steps 3–6 for all required target fields, then save the node’s configuration to apply changes.
Source: FastGPT official source
Applicability and version scope
Use this page for the documented Workflow nodes scenario. Confirm the FastGPT, dependency, API, and deployment versions in the official source before applying a change.
Safety guardrails
Use [REDACTED_CREDENTIAL] for credentials and private data. Confirm the documented environment and version before review.
Rollback guidance
Restore the prior technical-content authority snapshot. Restore saved configuration and data snapshots, then repeat the smallest verification scenario.