Newer Version Available

This content describes an older version of this product. View Latest

GenAiPlannerDefinition

Represents an agent planner service that uses a large language model (LLM) and a reasoning strategy to decompose a given task into smaller subtasks, identify the most suitable actions for each subtask, and invoke them. This object is available in API version 60.0 and later.

Supported SOAP API Calls

create(), delete(), describeSObjects(), query(), retrieve(), update(), upsert()

Supported REST API Methods

DELETE, GET, HEAD, PATCH, POST, Query

Special Access Rules

To access this object, Agents must be enabled in your org.

Fields

Field Details
Capabilities
Type
string
Properties
Filter, Group, Nillable, Sort
Description
A set of tags associated with the agent planner service definition.
Description
Type
textarea
Properties
Filter, Group, Sort
Description
A description explaining the general purpose and domain of the agent planner service definition.
DeveloperName
Type
string
Properties
Filter, Group, Sort
Description
The unique name for this object.
FullName
Type
string
Properties
Create, Group, Nillable
Description
The full name of the associated object in Metadata API. The full name can include a namespace prefix.
Language
Type
picklist
Properties
Defaulted on create, Filter, Group, Nillable, Restricted picklist, Sort
Description
The language of the GenAiPlannerDefinition. The value for this field is the language value of the org.
MasterLabel
Type
string
Properties
Filter, Group, Sort
Description
The master label of the agent planner service definition.
NamespacePrefix
Type
string
Properties
Filter, Group, Nillable, Sort
Description
The namespace of the GenAiPlannerDefinition.
PlannerType
Type
picklist
Properties
Filter, Group, Restricted picklist, Sort
Description
A particular approach to problem solving that is given as prompt instructions to a large language model (LLM).
Possible values are:
  • AiCopilot__ReAct—Uses a reactive planning strategy to solve problems with the LLM. This strategy consists of prompting the LLM to generate the next step in response to an event and the current context. It differs from a sequential planner in that it doesn’t plan more than one step ahead of time.
  • AiCopilot__SequentialPlannerIntentClassifier—Uses an intent classifier prompt and a sequential planner prompt. With each text input, the planner asks the LLM to generate a step-by-step plan to finish the goal. It plans first, then executes.