• Beta

    Define a LangGraph workflow using the entrypoint function.

    !!! warning "Beta" The Functional API is currently in beta and is subject to change.

    Function signature

    The wrapped function must accept at most two parameters. The first parameter is the input to the function. The second (optional) parameter is a LangGraphRunnableConfig object. If you wish to pass multiple parameters to the function, you can pass them as an object.

    Helper functions

    Streaming

    To write data to the "custom" stream, use the getWriter function, or the LangGraphRunnableConfig.writer property.

    State management

    The getPreviousState function can be used to access the previous state that was returned from the last invocation of the entrypoint on the same thread id.

    If you wish to save state other than the return value, you can use the entrypoint.final function.

    Type Parameters

    • InputT

      The type of input the entrypoint accepts

    • OutputT

      The type of output the entrypoint produces

    Parameters

    • optionsOrName: string | EntrypointOptions

      Either an EntrypointOptions object, or a string for the name of the entrypoint

    • func: OutputT extends AsyncGenerator<unknown, unknown, unknown>
          ? never
          : OutputT extends Generator<unknown, unknown, unknown>
              ? never
              : ((input, config) => OutputT)

      The function that executes this entrypoint

    Returns Pregel<Record<string, PregelNode<InputT, EntrypointReturnT<OutputT>>>, {
        __end__: LastValue<EntrypointReturnT<OutputT>>;
        __previous__: LastValue<EntrypointFinalSaveT<OutputT>>;
        __start__: EphemeralValue<InputT>;
    }, Record<string, unknown>, InputT, EntrypointReturnT<OutputT>>

    A Pregel instance that can be run to execute the workflow

    Example: Using entrypoint and tasks

    import { task, entrypoint } from "@langchain/langgraph";
    import { MemorySaver } from "@langchain/langgraph-checkpoint";
    import { interrupt, Command } from "@langchain/langgraph";

    const composeEssay = task("compose", async (topic: string) => {
    await new Promise(r => setTimeout(r, 1000)); // Simulate slow operation
    return `An essay about ${topic}`;
    });

    const reviewWorkflow = entrypoint({
    name: "review",
    checkpointer: new MemorySaver()
    }, async (topic: string) => {
    const essay = await composeEssay(topic);
    const humanReview = await interrupt({
    question: "Please provide a review",
    essay
    });
    return {
    essay,
    review: humanReview
    };
    });

    // Example configuration for the workflow
    const config = {
    configurable: {
    thread_id: "some_thread"
    }
    };

    // Topic for the essay
    const topic = "cats";

    // Stream the workflow to generate the essay and await human review
    for await (const result of reviewWorkflow.stream(topic, config)) {
    console.log(result);
    }

    // Example human review provided after the interrupt
    const humanReview = "This essay is great.";

    // Resume the workflow with the provided human review
    for await (const result of reviewWorkflow.stream(new Command({ resume: humanReview }), config)) {
    console.log(result);
    }

    Example: Accessing the previous return value

    import { entrypoint, getPreviousState } from "@langchain/langgraph";
    import { MemorySaver } from "@langchain/langgraph-checkpoint";

    const accumulator = entrypoint({
    name: "accumulator",
    checkpointer: new MemorySaver()
    }, async (input: string) => {
    const previous = getPreviousState<number>();
    return previous !== undefined ? `${previous } ${input}` : input;
    });

    const config = {
    configurable: {
    thread_id: "some_thread"
    }
    };
    await accumulator.invoke("hello", config); // returns "hello"
    await accumulator.invoke("world", config); // returns "hello world"

    Example: Using entrypoint.final to save a value

    import { entrypoint, getPreviousState } from "@langchain/langgraph";
    import { MemorySaver } from "@langchain/langgraph-checkpoint";

    const myWorkflow = entrypoint({
    name: "accumulator",
    checkpointer: new MemorySaver()
    }, async (num: number) => {
    const previous = getPreviousState<number>();

    // This will return the previous value to the caller, saving
    // 2 * num to the checkpoint, which will be used in the next invocation
    // for the `previous` parameter.
    return entrypoint.final({
    value: previous ?? 0,
    save: 2 * num
    });
    });

    const config = {
    configurable: {
    thread_id: "some_thread"
    }
    };

    await myWorkflow.invoke(3, config); // 0 (previous was undefined)
    await myWorkflow.invoke(1, config); // 6 (previous was 3 * 2 from the previous invocation)

Methods

Methods

  • Beta

    A helper utility for use with the functional API that returns a value to the caller, as well as a separate state value to persist to the checkpoint. This allows workflows to maintain state between runs while returning different values to the caller.

    !!! warning "Beta" The Functional API is currently in beta and is subject to change.

    Type Parameters

    • ValueT

      The type of the value to return to the caller

    • SaveT

      The type of the state to save to the checkpoint

    Parameters

    Returns EntrypointFinal<ValueT, SaveT>

    An object with the value and save properties

    Example

    return entrypoint.final({
    value: "result for caller",
    save: { counter: currentCount + 1 }
    });