Agent Setup
Describe a task, the rules the agent must follow, and the current context. Then click Run Agent to see how it behaves.
Task
What is the agent being asked to do?
Rules & Constraints
One rule per line. Use words like "must", "never", "cannot", or "avoid" for hard constraints.
Context & Signals
Observations, SLAs, incident pressure, or other signals the agent is aware of.
Behavior Summary
Agent has not been run yet.
Interpreted Task
The agent’s understanding of what it’s actually being asked to optimize.
Rules & Constraints Applied
- No rules processed yet.
Context & Drift Signals
- No context signals analyzed yet.
Decision & Rationale
Once the agent runs, this will describe its primary decision and why it chose that behavior.
Execution Trace
Run the agent to see a step-by-step trace of how rules and context were interpreted.