One CRD, any agent
A single Agent custom resource registers your container image with the
runtime. Routing, scaling, and pod lifecycle are handled — you just
bring the HTTP handler.
Concurrency-aware
The per-pod sidecar enforces in-flight caps and surfaces live load. Replicas auto-scale to keep up with traffic without exceeding the configured per-pod ceiling.
Streaming-native
SSE, chunked HTTP, and WebSocket upgrades pass through the gateway with immediate flushing. No buffering, no SSE tax.
Prometheus-native observability
Every component exposes krypton_* series — invocations, latency,
desired replicas, scaler decisions, sidecar in-flight. A starter
Grafana dashboard ships in the repo.
BYO ingress
The gateway ships as a ClusterIP. Put your existing ingress (Envoy / Nginx / ALB / Cloudflare) in front for TLS, auth, rate limiting — Krypton doesn’t reinvent any of it.
MCP, first-class
Run any HTTP-transport MCP server as an Agent, or wrap a stdio MCP
binary in the bundled bridge. The operator UI introspects each
server’s tools.