Connect an AI agent to your CubePath infrastructure with the Model Context Protocol (MCP). The CubePath MCP Server is hosted at https://mcp.cubepath.com, so there is nothing to install or run locally: point your MCP client at the endpoint.

https://mcp.cubepath.com/mcp

Connect without a token

Claude Code, the claude.ai connector and any client that implements the MCP authorization flow need no credential at all:

claude mcp add --transport http cubepath https://mcp.cubepath.com/mcp

A browser opens where you choose the organization the assistant may reach and tick the permissions you are granting. Nothing to create beforehand, nothing to paste, and nothing left in a config file.

Three things worth knowing about what you approve there:

  • The organization is fixed. The assistant reaches the one you picked and cannot move to another.
  • You can grant less than the application asked for. Untick anything you would rather it did not touch, including everything that writes.
  • It never exceeds your own role. Permissions are re-checked on every request against your current membership, so if your role changes, so does what the assistant can do.

Manage it afterwards from Account → Connected apps: you can see when each application last acted and disconnect any of them. Disconnecting takes effect on that application's next request, not when a token happens to expire.

The server speaks JSON-RPC 2.0 over streamable HTTP, one POST per message, and never accepts a session cookie - so no logged-in browser can drive it. Setting this up somewhere with no browser to open, like CI or a container? Use an API token as an X-API-Key header instead.

What your agent can do

150 tools cover your whole CubePath account, grouped by product:

  • Compute — VPS, baremetal, Kubernetes, availability groups.
  • Networking — load balancers, DNS, floating IPs, NAT gateways, private networks, tunnels, firewalls.
  • Delivery and data — CDN, managed databases, transcoder.
  • Platform — projects, SSH keys, alerts, status, DDoS.

Tools are named after what they do: cubepath_vps_create, cubepath_k8s_cluster_list, cubepath_dns_record_create.

Scopes decide what the agent sees

tools/list returns only the tools your token's scopes allow, so a read-only token exposes read-only tools and nothing else. Give each agent a token with the minimum scopes it needs.

Every tool is annotated as read-only or destructive, which lets your client ask you to confirm before something like cubepath_vps_destroy runs.

Long-running operations

Deploying a VPS or destroying a cluster is queued, and the tool returns straight away. Ask for the resource status to follow progress instead of calling the tool again: a second call queues a second operation.

Limits

  • 120 requests per minute per IP address on the MCP endpoint.
  • Request body up to 256 KB, and batches of up to 20 messages.
  • Tool results are capped at roughly 15 KB. Long listings are trimmed and the tool states what it left out, so a single response cannot fill the model's context.
  • GET returns 405. The server is stateless: no SSE, no sessions.