Implementation of a generative agent that can learn and form new memories over time. It extends the BaseChain class, which is a generic sequence of calls to components, including other chains.

Example

const tommie: GenerativeAgent = new GenerativeAgent(
new OpenAI({ temperature: 0.9, maxTokens: 1500 }),
new GenerativeAgentMemory(
new ChatOpenAI(),
new TimeWeightedVectorStoreRetriever({
vectorStore: new MemoryVectorStore(new OpenAIEmbeddings()),
otherScoreKeys: ["importance"],
k: 15,
}),
{ reflectionThreshold: 8 },
),
{
name: "Tommie",
age: 25,
traits: "anxious, likes design, talkative",
status: "looking for a job",
},
);

await tommie.addMemory(
"Tommie remembers his dog, Bruno, from when he was a kid",
new Date(),
);
const summary = await tommie.getSummary({ forceRefresh: true });
const response = await tommie.generateDialogueResponse(
"USER says Hello Tommie, how are you today?",
);

Hierarchy

Constructors

Properties

longTermMemory: GenerativeAgentMemory
name: string
status: string
traits: string
verbose: boolean

Whether to print out response text.

age?: number
callbacks?: Callbacks
memory?: BaseMemory
metadata?: Record<string, unknown>
tags?: string[]

Accessors

  • get inputKeys(): string[]
  • Returns string[]

  • get outputKeys(): string[]
  • Returns string[]

Methods

  • Adds a memory to the agent's long-term memory.

    Parameters

    • memoryContent: string

      The content of the memory to add.

    • Optional now: Date

      Optional current date.

    • Optional metadata: Record<string, unknown>

      Optional metadata for the memory.

    • Optional callbacks: Callbacks

      Optional Callbacks instance.

    Returns Promise<ChainValues>

    The result of adding the memory to the agent's long-term memory.

  • Run the core logic of this chain and add to output if desired.

    Wraps _call and handles memory.

    Parameters

    Returns Promise<ChainValues>

  • Creates a new LLMChain with the given prompt and the agent's language model, verbosity, output key, and memory.

    Parameters

    • prompt: PromptTemplate<any, any>

      The prompt to use for the LLMChain.

    Returns LLMChain<string, LLMType>

    A new LLMChain instance.

  • Computes the agent's summary by summarizing the agent's core characteristics given the agent's relevant memories.

    Parameters

    Returns Promise<string>

    The computed summary as a string.

  • Generates a dialogue response to the given observation.

    Parameters

    • observation: string

      The observation to generate a dialogue response for.

    • Optional now: Date

      Optional current date.

    Returns Promise<[boolean, string]>

    A boolean indicating whether to continue the dialogue and the output string.

  • Generates a reaction to the given observation.

    Parameters

    • observation: string

      The observation to generate a reaction for.

    • Optional now: Date

      Optional current date.

    Returns Promise<[boolean, string]>

    A boolean indicating whether to continue the dialogue and the output string.

  • Extracts the action of the given entity from the given observation.

    Parameters

    • observation: string

      The observation to extract the action from.

    • entityName: string

      The name of the entity to extract the action for.

    • Optional runManager: CallbackManagerForChainRun

      Optional CallbackManagerForChainRun instance.

    Returns Promise<string>

    The extracted action as a string.

  • Extracts the observed entity from the given observation.

    Parameters

    • observation: string

      The observation to extract the entity from.

    • Optional runManager: CallbackManagerForChainRun

      Optional CallbackManagerForChainRun instance.

    Returns Promise<string>

    The extracted entity as a string.

  • Returns a full header of the agent's status, summary, and current time.

    Parameters

    • config: {
          forceRefresh?: boolean;
          now?: Date;
      } = {}

      Optional configuration object with current date and a boolean to force refresh.

      • Optional forceRefresh?: boolean
      • Optional now?: Date

    Returns string

    The full header as a string.

  • Gets the agent's summary, which includes the agent's name, age, traits, and a summary of the agent's core characteristics. The summary is updated periodically through probing the agent's memories.

    Parameters

    • Optional config: {
          forceRefresh?: boolean;
          now?: Date;
      }

      Optional configuration object with current date and a boolean to force refresh.

      • Optional forceRefresh?: boolean
      • Optional now?: Date
    • Optional runManager: CallbackManagerForChainRun

      Optional CallbackManagerForChainRun instance.

    Returns Promise<string>

    The agent's summary as a string.

  • Invoke the chain with the provided input and returns the output.

    Parameters

    Returns Promise<ChainValues>

    Promise that resolves with the output of the chain run.

  • Parses a newline-separated string into a list of strings.

    Parameters

    • text: string

      The string to parse.

    Returns string[]

    An array of strings parsed from the input text.

  • Create a new runnable sequence that runs each individual runnable in series, piping the output of one runnable into another runnable or runnable-like.

    Type Parameters

    • NewRunOutput

    Parameters

    • coerceable: RunnableLike<ChainValues, NewRunOutput>

      A runnable, function, or object whose values are functions or runnables.

    Returns RunnableSequence<ChainValues, Exclude<NewRunOutput, Error>>

    A new runnable sequence.

  • Parameters

    Returns Promise<string>

  • Stream output in chunks.

    Parameters

    Returns Promise<IterableReadableStream<ChainValues>>

    A readable stream that is also an iterable.

  • Stream all output from a runnable, as reported to the callback system. This includes all inner runs of LLMs, Retrievers, Tools, etc. Output is streamed as Log objects, which include a list of jsonpatch ops that describe how the state of the run has changed in each step, and the final state of the run. The jsonpatch ops can be applied in order to construct state.

    Parameters

    Returns AsyncGenerator<RunLogPatch, any, unknown>

  • Summarizes memories that are most relevant to an observation.

    Parameters

    • observation: string

      The observation to summarize related memories for.

    • Optional runManager: CallbackManagerForChainRun

      Optional CallbackManagerForChainRun instance.

    Returns Promise<string>

    The summarized memories as a string.

  • Default implementation of transform, which buffers input and then calls stream. Subclasses should override this method if they can start producing output while input is still being generated.

    Parameters

    Returns AsyncGenerator<ChainValues, any, unknown>

Generated using TypeDoc