What is an Agent?
An agent is an autonomous AI entity powered by a large language model. When you send a message to an agent, it:- Reasons about the task using its behavior prompt and conversation history
- Calls tools if needed (built-in functions, integrations, or other agents)
- Iterates until it has a final answer or reaches the step limit
- Returns a final output with full execution trace
Agent lifecycle
Agent statuses
| Status | Description |
|---|---|
| Active | Agent is available for execution |
| Inactive | Agent is disabled — cannot be run |
| Archived | Agent is soft-deleted |
Run statuses
| Status | Description |
|---|---|
| Pending | Run created, waiting for execution |
| Running | Agent is actively processing |
| Completed | Agent finished successfully |
| Failed | Execution encountered an error |
| Cancelled | Run was cancelled by user or system |
Quality assurance
When enabled, a QualityChecker system agent evaluates each run’s output:- Quality Score — 0.0 to 1.0 rating
- Verdict —
pass,warn, orfail - Reasoning — explanation of the assessment
- Issues — specific problems found (if any)