Style & Behavior
The Style & Behavior section defines the visual appearance, tone, and operational behavior of your AI agent. These settings apply to how the agent appears in your product, how it communicates, and how it handles conversations.
1. Style
Agent Name (Optional)
Defines the display name shown to end users in the chat header.
Primary Color
Sets the main color used in the chat header and message bubbles.
Agent Icon
Upload or select an icon to represent the agent in the chat interface.
Position
Selects where the agent appears on the screen (e.g., bottom right, middle right, etc.).
Border Animation
Toggles an animated border effect around the agent bubble.
Border Colors
Defines the colors used for the animated or static border.
Prompt Hint
A short text shown in the input placeholder to guide users on how to start (e.g., “Ask me anything about your workflows…”).
Disclaimer (PRO)
Displays a “Powered by Foldspace” label.
In Pro accounts, this can be customized or removed.
Chat Font
Selects the font used in the chat interface.
Font Size
Adjusts the text size for chat messages.
Advanced Configuration
Expands additional style settings if available.
2. Behavior
Role
Defines the agent’s functional role (e.g., Support, Sales, or Custom).
This determines tone defaults and how some features behave.
Tone
Selects the tone used for responses (Friendly, Neutral, or Professional).
Elaboration Level
Controls how detailed responses are (Concise, Balanced, or Detailed).
Reasoning Level
Determines how much reasoning depth the model uses when forming answers (Low, Medium, High).
Allow Using Public Information
If enabled, the agent may reference public data sources when responding.
Allow Addressing Competitive Inquiries
Allows the agent to handle questions related to competitors (e.g., comparisons).
Enable Screen Sharing (formerly Page Reader Data)
Grants the agent structured access to the visible page content to understand user context and provide relevant responses or actions.
General
Offer Support Handoff
When enabled, the agent can forward a conversation to an external support channel.
Forwarding destinations are configured under Integrations (email, HubSpot, webhook, or Zendesk).
Slack is not used for customer handoff—it’s reserved for internal agents.
Customize Error Response
Defines the fallback message when no valid answer is available.
Instructions
Optional field for additional behavioral rules or context (e.g., “Keep responses short and friendly”).
Product Name
The name of the product the agent supports.
Product Description
Provides product context to improve the accuracy of answers and suggestions.
3. Starters
Enable Starters
Toggles the display of quick-start prompts that appear when the chat opens.
Hint
Short text displayed above the starter prompts (e.g., “Need help? Here are some topics we can start with.”).
Categories
- Knowledge: Starters that retrieve information from the knowledge base.
- Action: Starters that trigger defined system actions or workflows.
The leftmost category is used as the default when the widget is fully expanded.
Categories also determine the visual sorting order from left to right in the compact view.
4. Voice
Speech to Text
Enables input via microphone transcription.
Voice Chat (PRO)
Activates two-way voice interaction (available in Pro plans).
Voice Chat Type
Selects the voice model used for responses (e.g., Alloy).
Voice Speed
Adjusts the playback speed for spoken responses.
Voice Instructions
Defines additional guidelines for voice interactions.
This includes how to:
- Handle tone and speaking style
- Refer to actions or tools verbally
- Summarize data briefly rather than reading text in full
✅ With Style & Behavior, your agent becomes more than functional — it feels like a natural part of your product, speaking in your brand’s voice and guiding users the way you want.
Updated about 1 month ago