Agents
Configuration Options

Configuration Options

Detailed guide to all agent configuration options.

Basic Settings

Name and Description

  • Name — Identifies the agent in your dashboard
  • Description — Internal notes about the agent's purpose

System Prompt

The instructions that control agent behavior. See System Prompts for best practices.

Response Configuration

Temperature

Controls how creative or focused the agent's responses are:

ValueBehavior
0.0 - 0.3Focused, deterministic responses
0.4 - 0.7Balanced creativity and consistency
0.8 - 1.0More creative, varied responses

Recommendation: Start with 0.3-0.5 for factual Q&A based on documents.

Max Response Length

Limits how long responses can be. Consider:

  • Short (256 tokens) — Quick answers, mobile-friendly
  • Medium (512 tokens) — Standard responses
  • Long (1024+ tokens) — Detailed explanations

Knowledge Settings

Document Access

Control which documents your agent can search:

  • All documents — Search entire knowledge base
  • Selected folders — Limit to specific collections

Retrieval Settings

Fine-tune how the agent finds information:

  • Number of sources — How many document chunks to retrieve
  • Similarity threshold — Minimum relevance score

Conversation Settings

Greeting Message

The message displayed when a conversation starts. Leave empty for no greeting.

Conversation Memory

How the agent handles conversation context:

  • Full context — Remembers entire conversation
  • Limited context — Only recent messages

Advanced Settings

Model Selection

Choose the underlying AI model (if available on your plan).

Custom Instructions

Additional instructions that supplement the system prompt.

Changes to agent settings take effect immediately. Test in the Playground after making changes.