Response Generation controls how your agent formulates and delivers replies — the output format, the model that writes the answer, the response language, and the tone, style, and format of the prose.
Overview
The section's own description: "Set up how your agent formulates and delivers responses, including communication style and output format." From here you choose the output format, the model that writes the answer, the language behavior, and the tone, style, and format of the prose, plus options for showing the agent's reasoning, using graph-based retrieval, and watermarking generated images. The section saves automatically as you change it. These same settings appear for AI teams as well as individual agents.
Output Format
Use Output Format to choose how the agent returns its answer:
Option | Description |
|---|---|
Text | A normal written reply. The default. |
Object | Structured data. Opens a schema builder where you define the fields the agent must return, so the output is machine-readable and consistent. |
Pick Text for conversational answers and Object when another system consumes the agent's output and needs a fixed shape. The Answer Generation Model picker beneath it sets which model writes the response — "Creates relevant answers based on selected AI model."
Answer Generation Model, Language Type, Tone, Style and Format apply only to Text output — they are hidden when you choose Object.
In the schema builder, each field has a type — String, Number, Integer, Boolean, Object, or Array — and an Array field's items can themselves be edited recursively (including nested objects), so you can describe multi-level structured output. You can also generate a starting schema from your instructions, or paste one in via the JSON import option.
Note: Clear All needs a second confirming click (the button turns red and relabels to Confirm) before it wipes the schema. Once cleared, the only way back is to Cancel the editor without saving. Editing the schema again after arming it, in either the visual or JSON view, disarms the confirmation back to a plain Clear All.
Response Language
Use Language Type to choose how the agent picks a language:
User Locale — "Responses match the user's locale settings." (the default)
Same as User Query — "Responses use the same language as the query."
Fixed Response Language — "Chosen language will be used for every response." Selecting this reveals a Select Language picker.
Use User Locale or Same as User Query for multilingual audiences; use Fixed when every reply must be in one specific language regardless of who's asking.
Tone, Style, and Format
Three pickers shape the writing:
Picker | Options |
|---|---|
Tone | Formal, Informal, Friendly, Empathetic, Sarcastic, Enthusiastic, Neutral, or Optimistic (default Neutral). |
Style | Conversational, Technical, Narrative, Analytical, Descriptive, Instructional, or Minimalist (default Conversational). |
Format | Paragraph, Bullet points, Outline, Normal, or Markdown (default Paragraph). |
Match these to your audience — for example Friendly + Conversational + Paragraph for end-customer support, or Neutral + Technical + Bullet points for an internal IT agent.
Chain of Thought, Graph-Based Retrieval, and Watermarking
Three toggles add advanced behavior:
Chain of Thought — "Show the agent's step-by-step reasoning process" (on by default). Turn it off to hide the reasoning and return only the answer.
Graph-Based Retrieval — "Use graph traversal and relationship analysis to find contextually relevant answers" (off by default). Useful when answers depend on relationships between facts, not just matching text.
Watermarking — "Add subtle identifiers to Agent-generated images for verification" (off by default). When on, you choose a Watermarking Technique (Frequency Based) and optional Watermarking Text.