<aside> 🧭

Navigation:

• Getting Started

Set Up an Organization

• Developer Guide
• Agent Configuration
• Prompt Engineering

LLM Models

Model Evaluation Tool

Logs

Contact us

</aside>

The Logs section in the Deploy.AI Admin Dashboard gives you full visibility into your agents' activity, making it easier to debug issues, monitor usage, and track system behavior over time.

What You Can See in Logs

Logs display detailed records of how agents interact with users and systems. Each log entry typically includes:

• User messages sent to the agent
• Agent responses, including tool usage or function calls
• Tool activations, like RAG searches or API integrations
• Errors or system events captured during runtime
• Timestamps for every request/response cycle
• User identifiers to track interactions

Logs Modes

There are two views to choose from:

• Messages: A user-friendly chat-like timeline of the conversation between a user and the agent.
• Advanced Logs: A structured JSON view of all low-level request and response data, perfect for debugging tools, inspecting prompt input/output, or diagnosing errors.

When to Use Logs

Use logs to understand and fine-tune agent behavior. They help identify why a response was generated, whether tools triggered correctly, and how prompts were interpreted. Logs also reveal user interaction patterns and make it easier to track down and resolve errors.