Self-hosted docs
Automation
Automation should reduce risk, not hide it. Every action remains policy-bound and auditable.
SLA-aware reroute
When availability or latency degrades, traffic can move to approved fallback endpoints automatically.
- Example: Latency > 2000ms triggers failover to approved fallback endpoint
- Example: Availability < 99.9% for 5 minutes triggers automatic reroute
- Example: Fallback endpoints must pass policy checks before activation
- Request: POST /v1/failover/trigger {route_id: "prod-chat", reason: "latency_degradation", fallback_endpoint: "backup-chat"}
- Response: {status: "activated", original_endpoint: "primary-chat", fallback_endpoint: "backup-chat", timestamp: "2024-03-15T14:35:00Z"}
- API Reference: See [Failover Trigger](/docs/api#failover-trigger) for full contract details
Containment playbooks
Trigger route pause, policy tightening, and delivery workflows when correlated risk thresholds are crossed.
- Example: Correlated incidents trigger route pause and notification workflow
- Example: Policy tightening reduces token limits and increases safety thresholds
- Example: Delivery workflows route alerts to appropriate teams based on severity
- Request: POST /v1/containment/activate {incident_id: "inc-789", action: "pause_route", severity: "high"}
- Response: {containment_id: "con-101", status: "active", actions: ["route_paused", "alerts_sent"]}
- API Reference: See [Containment Activation](/docs/api#containment-activation) for full contract details
Automation lifecycle examples
End-to-end workflows showing automation from trigger to completion.
- Example: Complete failover lifecycle with policy validation and rollback
- Trigger: POST /v1/automation/trigger {type: "failover", route_id: "prod-chat", reason: "latency_spike"}
- Validation: GET /v1/automation/validate {automation_id: "auto-123"} → {status: "policy_approved"}
- Execution: POST /v1/automation/execute {automation_id: "auto-123"} → {status: "completed", result: "traffic_rerouted"}
- Rollback: POST /v1/automation/rollback {automation_id: "auto-123"} → {status: "original_route_restored"}
- Audit: GET /v1/automation/audit {automation_id: "auto-123"} → {timeline: [...], decisions: [...]}