ActionPlannerOptions interface

Package:

@microsoft/teams-ai

Options used to configure an ActionPlanner instance.

Properties

defaultPrompt	The default prompt to use.
enableFeedbackLoop	If true, the feedback loop will be enabled for streaming responses.
endStreamHandler	Optional handler to run when a stream is about to conclude.
feedbackLoopType	The feedback loop type.
logRepairs	If true, repair attempts will be logged to the console.
max_repair_attempts	Maximum number of repair attempts to make.
model	Model instance to use.
prompts	Prompt manager used to manage prompts.
startStreamingMessage	Optional message to send a client at the start of a streaming response.
tokenizer	Optional tokenizer to use.

Property Details

defaultPrompt

The default prompt to use.

defaultPrompt: string | ActionPlannerPromptFactory<TState>

Property Value

string | ActionPlannerPromptFactory<TState>

Remarks

This can either be the name of a prompt template or a function that returns a prompt template.

enableFeedbackLoop

If true, the feedback loop will be enabled for streaming responses.

enableFeedbackLoop?: boolean

Property Value

boolean

endStreamHandler

Optional handler to run when a stream is about to conclude.

endStreamHandler?: PromptCompletionModelResponseReceivedEvent

Property Value

PromptCompletionModelResponseReceivedEvent

feedbackLoopType

The feedback loop type.

feedbackLoopType?: "default" | "custom"

Property Value

"default" | "custom"

logRepairs

If true, repair attempts will be logged to the console.

logRepairs?: boolean

Property Value

boolean

Remarks

The default value is false.

max_repair_attempts

Maximum number of repair attempts to make.

max_repair_attempts?: number

Property Value

number

Remarks

The ActionPlanner uses validators and a feedback loop to repair invalid responses returned by the model. This value controls the maximum number of repair attempts that will be made before returning an error. The default value is 3.

model

Model instance to use.

model: PromptCompletionModel

Property Value

PromptCompletionModel

prompts

Prompt manager used to manage prompts.

prompts: PromptManager

Property Value

PromptManager

startStreamingMessage

Optional message to send a client at the start of a streaming response.

startStreamingMessage?: string

Property Value

string

tokenizer

Optional tokenizer to use.

tokenizer?: Tokenizer

Property Value

Tokenizer

Remarks

If not specified, a new GPTTokenizer instance will be created.