Try
Reference

ResponseNode

Final output node for chatbot responses. This node formats and returns the final response to the user.

Overview

The ResponseNode is a required exit point node that serves as the final output for all workflows. It accepts processed data from other nodes and formats it as the final response.

Note: Every workflow must include at least one ResponseNode. This is the only way to output final responses to users.

Inputs

input_data *

Type: string | Required: Yes

Combined input data from multiple connections. This can be connected from LanguageModelNode's response output, or from any other node that produces text output. If multiple nodes are connected, their outputs are automatically combined.

Outputs

final_response

Type: string

The final processed response that will be displayed to the user. This is the output that gets returned from the workflow execution.

Parameters

ResponseNode has no configurable parameters. It simply passes through the input data as the final response.

Multiple Input Connections

ResponseNode can accept multiple input connections. When multiple nodes are connected to ResponseNode, their outputs are automatically combined:

If you connect both a LanguageModelNode and a KnowledgeBaseRetrievalNode to ResponseNode, the outputs will be combined as:

javascript
"Combined inputs: [LanguageModelNode output] [KnowledgeBaseRetrievalNode output]"

Tools Used

ResponseNode does not use any tools. It simply formats and returns the input data.

Example Usage

Here's how ResponseNode is typically used in workflows:

javascript
// Simple workflow:
QueryNode (query: "Hello")
  ↓ [query]
LanguageModelNode
  ↓ [response]
ResponseNode
  → Final output: "Hello! How can I help you today?"
javascript
// Multiple inputs:
QueryNode (query: "What is ConvoFlow?")
  ↓ [query]
LanguageModelNode
  ↓ [response]
ResponseNode ← Also receives input from DebugNode
  → Final output: Combined response

Error Handling

ResponseNode automatically detects error patterns in input data and handles them gracefully:

  • If input contains error prefixes like "Error:", "ERROR:", or "error:", it returns an error message
  • If input contains error phrases like "failed to", "not configured", "not found", it flags as error
  • Error responses are formatted as: "An error occurred in the workflow. Please check the error details."

Styling

ResponseNode has a distinctive pill-shaped design with a cyan border to indicate it's an end point:

  • Shape: Pill (rounded ends)
  • Border Color: Cyan (#06b6d4)
  • Background: Dark (#1f1f1f)
  • Subtitle: "END POINT"
  • Output Handles: Hidden (terminal node)
  • Icon: Send icon (📤)

Related Nodes

Was this page helpful?