Advanced Agent Settings
Most teams never need this tab - the defaults are tuned to work well, and a strong personality description does most of the shaping. Reach for the Advanced tab once you have watched a few real conversations and want to nudge a specific behavior. Everything here is optional, and anything you leave blank falls back to the platform default.
Open it from Agents > your agent > the Advanced tab.
The agent Advanced tab showing the time limit, creativity and answer length settings, and the full instructions field
Answer Quality
Answer quality (also called smart mode) sets how much reasoning power this agent brings to each customer conversation:
- Best reasoning - for complex, multi-step, or sensitive conversations. The most capable model.
- Balanced - the most efficient choice for everyday support. The default.
- Fast - the quickest replies for simple questions and lookups.
Each option names the model it runs on. The choice applies to this agent only, so you can run a billing agent on Best reasoning and a store-hours agent on Fast. Higher gives better answers and uses more of your plan. Higher levels may be part of a paid plan - a locked option tells you which plan unlocks it. Want to run your agent on your own model provider and API key? Contact us about bring-your-own-key.
Steps and Time Limit
Your agent can take several steps in a single turn - search your Knowledge Base, call a tool, then write its answer. These settings cap how far it goes, to protect speed and cost:
- Steps it may take - the hard cap on tool-call rounds per turn (platform default: 50). Raise it if your agent gives up before finishing multi-step tasks; lower it for faster, simpler replies.
- Time limit (in seconds) - end a turn if it runs longer than this.
- Steps it may take also stops the agent if it repeats the same action too many times.
If you are not sure, leave these blank - the defaults suit normal support.
Answer Style
These apply to every turn:
- Creativity (also called temperature or randomness) - how creative the wording is, from 0 (same answer every time) to 2 (creative); the default 0.7 is balanced.
- Longest answer (also called max output tokens) - the cap on how long a reply can be (default 4096). Raise it if answers get cut off; lower it to keep replies brief.
- Think longer before answering - lets the model show its reasoning on complex decisions (Anthropic models only).
For support, lean toward lower Creativity. Customers want a correct, consistent answer far more than a creative one.
Instructions
The most powerful setting here splits your agent's Full instructions (the instructions it always follows, also called the system prompt) into layers:
Your agent's full instructions are assembled in layers: its personality (the Full instructions field on the Personality tab), the platform's built-in conduct rules, and the Domain instructions you write here - no layer replaces another.
- Domain instructions - what the agent does: workflow, rules, and handover triggers (for example: "Confirm the order number before discussing a refund; if a claim is over $150, hand over to a person"). This is added as its own section on top of everything else.
Most teams only need one or two firm rules here. A See example patterns link offers ready-made starting points. Guidance tied to a specific skill belongs on the skill itself (the Skills tab) - it travels with the skill and disappears when the skill is removed.
These settings change how your agent behaves, so test after editing. Use the preview panel to run a few real questions - including ones your agent should refuse or hand over - and confirm it still responds the way you expect before you turn it on.
What's Next
With your agent configured the way you want, it is time to put it on your site. Head to Add Your Agent to Your Website for the steps for your website's platform.
Was this page helpful?