Manager
A Kubernetes-native control plane to install the stack, reconcile MCPServer resources, and keep deployments healthy.
Spec 00 · MCP infrastructure for Kubernetes · Alpha
Deploy, govern, and broker MCP servers using a Kubernetes-native control plane.
MCP Runtime gives platform teams one place to deploy, route, govern, and observe Model Context Protocol servers. Define services once, enforce policy on the live request path, and give AI agents, IDE users, and direct human workflows a controlled way to reach internal tools.
The public entrypoint at platform.mcpruntime.org is not a public or private marketplace. It is a live preview of the control surface companies get when they deploy MCP Runtime in their own cluster: easy deployment, governed access, brokered execution, audit, compliance evidence, and operations on infrastructure they control.
01 / Definition
One manifest becomes a governed MCP endpoint.
apiVersion: mcpruntime.org/v1alpha1
kind: MCPServer
metadata:
name: payments
spec:
image: registry.example.com/payments-mcp:v1.0.0
port: 8088
publicPathPrefix: payments
gateway:
enabled: true
analytics:
enabled: trueShip internal MCP servers with the same discipline as any other production service.
- Define the server, image, port, route, gateway, and analytics in one resource.
- Let the operator reconcile services, ingress, rollouts, and updates.
- Keep policy, audit, and analytics attached to the endpoint from day one.
02 / Architecture
The pieces companies need to run MCP safely.
Registry
Image and deployment workflows for publishing MCP servers into environments your platform team controls.
Broker
A governed gateway that brokers agent access, applies policy before tool calls run, and records what happened.
03 / Capabilities
Built for enterprise MCP operations.
MCP Runtime turns scattered MCP experiments into one operating surface for platform teams: deploy servers, apply access policy, broker requests, and inspect audit and traffic data from one place.
Managed deployments
Define MCP servers once as Kubernetes CRDs. The operator handles services, rollouts, routes, and updates in a GitOps-friendly workflow.
Company access model
Grants and sessions make consent, trust level, expiry, and revocation inspectable Kubernetes resources so governed access is the default path.
Hosted platform shape
The public platform.mcpruntime.org experience shows how the platform looks after deployment. It is a product preview, not a vendor storefront, listing site, or MCP marketplace.
Brokered tool calls
Route agent traffic through a gateway that can enforce policy before sensitive tools are executed.
Operational visibility
Observe MCP calls with policy decisions, timing, tool names, status, and dashboard-ready event data instead of treating agent traffic like unstructured logs.
Audit-ready history
Every governed request is recorded for security review, incident response, and compliance workflows.
Platform stack included
Registry, ingress, API, ingest, processor, UI, Grafana, and Prometheus are wired for local evaluation.
04 / Position
More than discovery.
Public MCP directories such as Glama, Smithery, Docker MCP Catalog, PulseMCP, mcp.so, and client-specific catalogs help people discover or install MCP servers. MCP Runtime is different: it is an open-source runtime platform for companies that need to host, deploy, govern, audit, and operate MCP servers inside their own Kubernetes environment.
Deployment first
Define MCP servers as Kubernetes resources and let the platform handle routes, rollouts, registry wiring, and status.
Open source runtime
As of April 2026, we have not found another open-source MCP product that combines a deployable Kubernetes control plane, registry workflow, brokered request path, access/session model, audit pipeline, and operational control surface.
Governed by design
Companies can manage MCP access for agents, IDEs, and humans with grants, sessions, policy decisions, and audit evidence.
05 / Runbook
Install the stack, then connect servers.
-
Build
go build -o bin/mcp-runtime ./cmd/mcp-runtime -
Bootstrap
./bin/mcp-runtime bootstrap -
Install
./bin/mcp-runtime setup --test-mode -
Verify
./bin/mcp-runtime status
06 / Use Cases
Where companies put it to work.
Agent tool access
Expose internal tools to agents with controlled access, request inspection, and audit trails.
Internal MCP operations
Give teams a shared control surface to publish, operate, govern, and inspect MCP servers instead of one-off sidecars.
Private platform preview
Use the public platform to understand the deployed product, then run the same control surface inside a company cluster for owned MCP servers.
Cross-team governance
Grant temporary, reviewable access between teams with clear trust ceilings and easy revocation.
Compliance review
Give security and compliance teams searchable request history instead of scattered service logs.
07 / Targets
Runs where your company already runs Kubernetes.
Kubernetes 1.25+. Self-hosted. Open source. Vendor-neutral across Kubernetes distributions. No proprietary control plane.
k3s
kind
minikube
kubeadm
EKS
GKE
AKS
OpenShift
Preview the platform, then run it yourself.
Open the public platform to see the control surface. Install MCP Runtime to give your company that same experience on private infrastructure, with governance and audit built into the runtime path.