Plugin Node belongs to Information Processing Node, supporting users to refer to official and custom plugins (for plugin creation method, please see Creating Based on API and Code).
In the Tencent Cloud Agent Development Platform (Tencent Cloud ADP), Tool means an API to complete specific scenario requirements or tasks, while Plugin refers to a collection of tools, meaning a plugin may contain one or more tools.
When creating a new plugin node in the canvas, system automatically opens a plugin selection dialog where you can choose required tools. Select it, then use the plugin node in the canvas.
Directions
Input variables
It means the parameters used when calling the plugin, supporting two ways: reference to ancestor node output variables and input fixed values. User customization is not supported.
Configuration
Description
Variable Name
The variable name can only contain letters, digits, or underscores, must start with a letter or underscore. Required.
Description
Description of this variable. Optional
Data Source
The data source of this variable supports two options: "refer" and "input". "Refer" allows selecting output variables from all preceding nodes, while "input" supports manually filling in a fixed value. Required
Type
The data type of this variable cannot be selected and defaults to the variable type of "refer" or the string type of "input".
When configuring a plugin based on API creation, input variables are divided into three parts: header, query, and body, with the following meanings:
Configuration
Description
header
Used for transmitting metadata, usually related to the request itself rather than business data, commonly used for identity authentication
query
Pass parameters through the url, often used for filtering, sorting, or pagination.
body
Transmit actual data content
Output Variable
The output variable processed by this node includes the result of calling the tool as well as runtime Error info (data type: object, this field is empty during normal operation). Manual addition is not supported.
Handling error
Exception handling can be enabled manually (exception handling is off by default), supporting exception retry and exception handling method configuration. The configuration content is as follows:
Configuration
Description
Max Retry Attempts
Maximum number of times to rerun when the node is running exceptionally. If retries exceed the set number of times, consider that node call failed and execute the exception handling method below. Default is 3 times.
Retry Interval
Interval between each rerun, default is 1 second.
Handling Method
Support three types, including "Output Specific Content", "Execution Exception Flow", and "Interrupt Flow".
Output Variable
When the exception handling method is selected as "Output Specific Content", the output variable returned when retries exceed the maximum number.
When the exception handling method is selected as "Output Specific Content", the workflow will not be interrupted after an exception occurs, and the node will return the output variable and variable value set by the user in the output content after exception retry.
When the exception handling method is selected as "Execution Exception Flow", the workflow will not be interrupted after an exception occurs, and the node will execute the exception handling process customized by users after exception retry.
When the exception handling method is selected as "Interrupt Process", there are no more setting items. The workflow execution will be interrupted after an exception occurs (interrupting the process may cause workflow data loss. Perform backup before the next operation).
Application Example
Before planning the travel route, first use the plugin to query the weather based on the time and location provided by the user.
Weather plugin configuration as follows:
FAQs
1. What is the difference between MCP plugin and API plugin?
MCP (Model Context Protocol) is an open protocol designed for LLM applications, aiming to standardize interactions between LLMs and external data sources or tools (such as databases, APIs, computing services). It provides a unified API specification, enabling developers to quickly integrate various features (like search, calculation, API calls) into LLM applications without the need to adapt each tool separately. API plugins are components that expand system functions by directly calling specific API interfaces, typically requiring separate adaptation logic for different protocols (e.g., REST, WebSocket) and authentication methods (e.g., OAuth, API Key).
2. How to create a plugin?
On the workflow configuration page, click "Create plugin" at the plugin node, or click "Create plugin" in the plugin center to access the API plugin creation page. Fill in basic information, add tools (API), and verify tools (API) to create an API plugin. In the plugin center, click "Integration MCP plugin" to access the MCP plugin creation page, which supports manual configuration and import of JSON-formatted configurations.
