send.md

title	sidebar_position	slug
Send	3	/send

Langflow enhances its functionality with custom components like Send. This component sends messages to the specified session of a self hosted Streamlit application.

Prerequisites

1. Setting up a Streamlit App: Follow the guide Setting up a Streamlit App to set up a Streamlit application in your workspace.

Component Functionality

The Send component allows you to:

• Sends message to specific streamlit chat session.

Component Usage

To incorporate the Send component into a Langflow flow:

1. Add the Send component to your flow.
2. Configure the component by providing:
   • SessionId: The ID of the session you want to interact with.
   • Role: The role that will be used to send the message.
3. Connect the component to other nodes in your flow as needed.
4. Initiate the flow to begin retrieving the messages of session from a Streamlit server.

Code Block for the Send Component

from langflow.custom import Component
from langflow.schema.message import Message
from langflow.inputs import MessageTextInput, DropdownInput, NestedDictInput, IntInput, StrInput, BoolInput
from json import loads, dumps


class Send(Component):
    display_name = "Send"
    description = "Send a streamlit chat message (webhook)."
    icon = "Streamlit"

    inputs = [
        MessageTextInput(
            name="session_id",
            display_name="Session ID",
            info="Session ID of streamlit. For each browser tab that connects to the Streamlit server, a new session is created.",
            placeholder="Enter the session ID",
            value="",
            required=True
        ),
        DropdownInput(
            name="role",
            display_name="Role",
            options=["ai", "user"],
            info="Message sender role.",
            placeholder="Enter the message sender role.",
            value="ai",
            required=True
        ),
        MessageTextInput(
            name="message",
            display_name="Message",
            info="Content of the message that will be sent.",
            value="",
            required=True
        ),
        NestedDictInput(
            name="tweaks",
            display_name="Tweaks",
            info="The tweaks can be used to modify the flow's parameters and components.",
            value={},
            required=True,
            advanced=True
        ),
        IntInput(
            name="port",
            display_name="Port",
            info="Streamlit API Port",
            value=7881,
            required=True,
            advanced=True
        ),
        StrInput(
            name="hostname",
            display_name="hostname",
            info="IP or hostname of Streamlit API",
            value="localhost",
            required=True,
            advanced=True
        ),
        BoolInput(
            name="rerun",
            display_name="Rerun Flow",
            info="If True, run the current flow again",
            value=True,
            required=True,
            advanced=True
        )
    ]

    outputs = [
        Output(display_name="Text", name="text", method="text_response"),
    ]


    async def text_response(self) -> Message:
        import requests
        flow_id = self.vertex.graph.flow_id
        resp = requests.post(f"http://{self.hostname}:{self.port}/api/v1/sessions/{self.session_id}/messages", json={"role": self.role, "content": self.message})
        if resp.status_code == 200:
            self.rerun and await self.run_flow(
                inputs={"input_value": ""}, flow_id=flow_id, tweaks=self.tweaks
            )
            return Message(text=dumps(loads(resp.content)))
        else:
            raise Exception("Timeout exception")

Example Usage

Example of using the Send component in a Langflow flow:

In this example, the Send component receives a text message from the OpenAI component and the sessionId from the Listen component as inputs, then sends the message to the Streamlit chat session.

Troubleshooting

If you encounter any issues while using the Send component, consider the following:

• Ensure the provided session ID is correct and the Streamlit application is accessible.
• Ensure the provided role is set appropriately.
• Consult the Streamlit Developers APP Page for documentation updates.