Manage Infrastructure with AI
Last updated
Was this helpful?
Who this is for: DevOps engineers, SREs, platform engineers, and network engineers who audit IONOS CLOUD infrastructure without leaving their AI client or opening the DCD console.
Use the MCP server to audit, explore, and cross-reference your IONOS CLOUD infrastructure through natural language, without switching away from your AI client.
Note: All tools are read-only. You can inspect and query resources, but the MCP server cannot power servers on or off, resize resources, attach storage, or make any modifications. Use the DCD or IONOS CLOUD API for write operations.
Most Compute Engine tools are scoped to a specific data center. When asking about servers, volumes, or LANs, include the data center name or ID for best results.
Ask your AI assistant to list and describe your resources:
List all my data centers and their locations.
Show me the servers in data center
abc-123.
What volumes are attached to server
web-01?
The assistant calls the relevant tools (list_datacenters, list_servers, list_server_volumes) and returns the results in a readable format.
List all LANs in my data centers.
What are the firewall rules for server
db-01?
Do I have any IP blocks reserved that are not in use?
List all network load balancers and their forwarding rules.
What NAT gateways are configured in data center
prod?
List all security groups and their inbound rules.
Which servers are in security group
web-tier?
You can chain questions to navigate your infrastructure:
Find all servers in Frankfurt and list their attached volumes.
Which data centers have load balancers configured?
Note: Compute Engine tools are registered at startup by default, so they are immediately available. If you have opted into lazy mode with IONOS_MCP_LOAD_MODE=lazy, ask your AI assistant to load the Compute tools first. For more information, see Selective Tool Loading.
Last updated
Was this helpful?
Was this helpful?