Best Practices
Agent Prompt Engineering
Every ATLAS agent requires a carefully structured system prompt. Here's the recommended format:
identity: {
role: "Specific purpose",
scope: "Knowledge domain",
limitations: ["List", "of", "constraints"]
}
Behavioral Guidelines
behavior: {
tone: "Professional but accessible",
response_length: "Concise but complete",
interaction_style: "Collaborative"
}
Knowledge Integration
knowledge: {
primary_domains: ["List", "of", "domains"],
confidence_threshold: 0.9,
update_frequency: "real-time"
}
For more detailed information, please check out our API Reference sections.