Overview of FastGPT System Tool Plugins
FastGPT system tool plugins extend platform functionality by integrating custom logic and external services. These plugins are built exclusively using two official FastGPT packages: @fastgpt-plugin/cli and @fastgpt-plugin/sdk-factory. The CLI handles project scaffolding, while the SDK provides standardized structures to define plugin behavior and configurations.
Mandatory SDK Declaration Fields
All system tool plugins require five core declarations implemented via the SDK. The following table lists each required field and its purpose:
| Field Name | Purpose |
|---|---|
| manifest | Defines core plugin metadata such as name, version, and display details |
| inputSchema | Validates user-provided input parameters for the plugin |
| outputSchema | Defines the structure and format of the plugin's returned output |
| secretSchema | Specifies required secure configuration values for plugin execution |
| handler logic | Implements the core functional logic of the plugin |
Build and Packaging Workflow
The standard workflow for developing and distributing system tool plugins follows these structured steps:
- Initialize Project: Use the
@fastgpt-plugin/clito generate a project skeleton for either a single standalone tool or a full tool suite. - Configure Plugin Logic: Use the SDK to populate all mandatory declaration fields and write the custom handler logic that executes the plugin's core functions.
- Validate and Test: Run local testing and validation workflows to confirm the plugin adheres to schema requirements and functions correctly.
- Package for Distribution: Execute the provided build, check, and pack operations to compile the plugin into a distributable
.pkgfile, which can be deployed to FastGPT environments.
For full, step-by-step development instructions including code examples and troubleshooting, refer to the System Tool Development Guide.
Source: FastGPT official source
Applicability and version scope
Use this page for the documented Model guides 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.