Mission-control incident simulation
Demonstrate the full OpsPilot lifecycle: alert intake, agent reasoning, evidence accumulation, safety gate pause, human approval, recovery monitoring, and postmortem generation.
Mission lifecycle
Agent activation board
Agent states activate as the mission progresses.
Production checkout latency anomaly is received and normalized.
Severity, affected service, environment, and blast radius are classified.
Metrics, logs, deployment records, and runbook signals are correlated.
Likely cause is ranked against evidence lineage and confidence.
Production remediation policy is evaluated before any action is allowed.
Operator reviews recommendation, risk, rollback plan, and policy reason.
Approved rollback is executed with reversible production controls.
Post-action telemetry confirms latency and cache health recovery.
Final incident record, evidence chain, and learning summary are generated.
Mission event stream
A judge-friendly incident story with step-by-step operational evidence.
Evidence appearing now
p95 latency crossed the critical threshold at 2.8s.
Event log
Mission outcome package
Final artefacts produced by the incident command workflow.
Judge scenario checkpoints
What the reviewer should notice during the simulation.