Operations

Production-facing setup for ingress, observability, and troubleshooting.

Operations pages assume Krypton is installed and you are preparing a cluster for shared or production use.

Common tasks

Task	Page
Put TLS, auth, rate limiting, or external routing in front of Krypton	Bring your own ingress
Scrape runtime metrics and install the Prometheus Operator monitors	Observability
Debug pods, gateway calls, UI health, model startup, or MCP handshakes	Troubleshooting

Front Krypton with Gateway API, Nginx, or any L7 ingress — plus webhooks and non-Kubernetes notes.

Metrics and other signals from the runtime.

Common failure modes, fixes, and frequently-asked questions.