Chat Output
On this page
Purpose
Chat Output displays a Message in the Playground and can store the resulting message for the active session.
When to use it
- Show an Agent Core response in a chat flow.
- Display a Message, Data value, or DataFrame as a chat response.
- End an interactive Playground conversation.
How to use it
- Add Chat Input and the component that generates a response.
- Connect the response to Inputs.
- Place Chat Output at the end of the chat path.
- Run the flow in the Playground.
Configuration
| Setting | What it controls | Recommended starting point |
|---|---|---|
| Store Messages | Stores the output when a session ID is available. | Leave on for multi-turn chat. |
| Sender Type | Sets the chat role. | Keep the default Machine. |
| Sender Name | Sets the display name. | Keep the default AI. |
| Session ID | Selects the stored-message session. | Leave blank to use the current session. |
| Data Template | Appears in the editor but does not change current Data conversion. | Leave at the default {text}. |
| Basic Clean Data | Affects conversion of list inputs. | Leave on unless a list needs its raw representation. |
Expected result
Chat Output returns an Output Message and displays it in the Playground. Data is formatted as JSON and DataFrame input becomes a Markdown table. A connected Message keeps its other fields where possible, but Chat Output overwrites its text, sender, sender name, session ID, flow ID, and source.
Storage requires both Store Messages and a session ID. The component has no direct credential or credit behavior.
Reference details
For a list of inputs, Chat Output converts each compatible item and joins the results into one message. A single DataFrame does not use Basic Clean Data.
Related components
- Agent Core supplies a conversational response.
- Chat Input starts the Playground conversation.
- Guardrail can route an approved or flagged Message to chat.