User Prompt
The User Prompt node acts as the execution entry point for a workflow. It captures the initial user-provided text input and emits it as a string value to downstream nodes. Use this node as the first node in any workflow that responds to user messages through the chatbot interface.
How It Works
The User Prompt node activates when a user sends a message through the chatbot interface. The node captures the incoming text and stores it in the workflow state, making it accessible to all connected downstream nodes.
As a trigger node, it responds to external events rather than processing data from upstream nodes. Execution begins here and flows outward to connected nodes.
Configuration
This node has no configuration parameters. It automatically captures the incoming user message and stores it in the workflow state.
Output
| Property | Value |
|---|---|
| Type | String |
| Content | User's input message |
| Access | ${system.user_query} or configured variable name |
Execution Lifecycle
| Phase | Behavior |
|---|---|
| Pre-execution | Awaits incoming user message |
| Execution | Captures message content, stores in workflow state |
| Post-execution | Passes control to connected downstream nodes |
The node transitions from Not Run to Completed immediately upon receiving input.
Usage Example
Standard conversational workflow:
User Prompt → Large Language Model → Response Output
Retrieval-Augmented Generation (RAG) workflow:
User Prompt → Vector Search → Large Language Model → Response Output
The user's input flows from the trigger through processing stages to the final response.
Limitations
- Single entry point: One User Prompt node per workflow
- Text only: Captures text messages; file attachments require separate handling
- No preprocessing: Passes input without modification
- Session-scoped: Captured message is specific to the current chat session
Best Practices
- Position the User Prompt node as the first node in the workflow
- Connect directly to processing nodes with clear edge paths
- Add validation or conditional nodes downstream for varied input types
- Use descriptive output variable names for clarity in complex workflows
Related Nodes
- Large Language Model - Process user input with AI models
- IF Conditional - Route workflow based on input content
- Vector Search - Find relevant documents based on user query
- Response Output - Return results to the user