LEARNING

System configuration

Settings

Protection policy

Observescore >= 30
Rate limitscore >= 60
Challengescore >= 80
Blockscore >= 90
Revokescore = 100

Runtime status

Detection engineProtected
LLM explainerUnavailable — no API key configured

Demo data

Clears incidents, sessions, and detector state — keeps learned baselines. Overridden incidents and their audit trail are never deleted. For a full re-seed (including baselines), run make reset from the CLI.