Class AbstractTextToModelParsingChatAgent<T>Abstract

Protected contentMatcherProviders

Protected contentMatchers

Protected defaultContentFactory

Protected defaultLanguageModelPurpose

Protected languageModelRegistry

Protected logger

Protected promptService

Protected recordingService

Protected callLlm

• callLlm(languageModel, messages, tools, token): Promise<LanguageModelResponse>

• Parameters

  • languageModel: LanguageModel
  • messages: ChatMessage[]
  • tools: undefined | ToolRequest[]
  • token: CancellationToken

  Returns Promise<LanguageModelResponse>

Protected getLanguageModel

• getLanguageModel(languageModelPurpose): Promise<LanguageModel>

• Parameters

  • languageModelPurpose: string

  Returns Promise<LanguageModel>

Protected getLanguageModelSelector

• getLanguageModelSelector(languageModelPurpose): LanguageModelRequirement

• Parameters

  • languageModelPurpose: string

  Returns LanguageModelRequirement

Protected getLlmSettings

• getLlmSettings(): undefined | {
      [key: string]: unknown;
  }

• Returns undefined | {
      [key: string]: unknown;
  }

  the settings, such as temperature, to be used in all language model requests. Returns undefined by default.

Protected getMessages

• getMessages(model, includeResponseInProgress?): Promise<ChatMessage[]>

• Parameters

  • model: ChatModel
  • includeResponseInProgress: boolean = false

  Returns Promise<ChatMessage[]>

Protected Abstract getSystemMessageDescription

• getSystemMessageDescription(): Promise<undefined | SystemMessageDescription>

• Returns Promise<undefined | SystemMessageDescription>

Protected getTools

• getTools(request): undefined | ToolRequest[]

• Parameters

  • request: ChatRequestModel

  Returns undefined | ToolRequest[]

  the list of tools used by this agent, or undefined if none is needed.

Protected handleError

• handleError(request, error): void

• Parameters

  • request: ChatRequestModelImpl
  • error: Error

  Returns void

Protected initializeContentMatchers

• initializeContentMatchers(): void

• Returns void

Protected onResponseComplete

• onResponseComplete(request): Promise<void>

• Invoked after the response by the LLM completed successfully.

  The default implementation sets the state of the response to complete. Subclasses may override this method to perform additional actions or keep the response open for processing further requests.

  Parameters

  • request: ChatRequestModelImpl

  Returns Promise<void>

Protected parseContents

• parseContents(text, request): ChatResponseContent[]

• Parameters

  • text: string
  • request: ChatRequestModelImpl

  Returns ChatResponseContent[]

Protected selectLanguageModel

• selectLanguageModel(selector): Promise<LanguageModel>

• Parameters

  • selector: LanguageModelRequirement

  Returns Promise<LanguageModel>