Civo MCP. Manage K8s, instances, and networks with chat.

You've got full control over your whole cloud setup with this server. It lets your AI client manage everything in Civo—from provisioning entire Kubernetes clusters to tweaking network firewalls—all using plain English commands, no dashboard clicking required.

Provisioning and Managing Compute Instances
You can launch a new virtual machine instance by calling create_instance, specifying the size and region. If you need more juice later on, use resize_instance to adjust the CPU or RAM. You'll also need to know what's available; check out list_sizes for current compute specs and list_disk_images to see templates.

Need a quick restart? Use soft_reboot_instance for a graceful OS reboot, or hit reboot_instance if you gotta hard reboot it immediately. You can power up an instance that's been shut down using start_instance, and remember to run stop_instance when you don't need it to save cash. For organization, tag your machines with retag_instance.

If you’re done with a machine but want to keep the credentials handy, upload new access keys via upload_ssh_key, or see what keys are already there by running list_ssh_keys.

Kubernetes Cluster Control
Managing K3s clusters is straightforward. You can build and provision an entirely new Kubernetes cluster environment using create_cluster. If you need to check the status of your setups, run list_clusters. To refresh a node's state without taking the whole cluster down, use recycle_cluster_node, and if you’re only listing what exists, list_clusters gets everything.

Networking and Security Configuration
Building out your network infrastructure is simple. First, create a brand new private network segment with create_network. When you need to lock down traffic, establish an isolated boundary using create_firewall, then use create_firewall_rule to add specific rules—like allowing port 80—or check existing security settings with list_firewall_rules. For domain names and DNS records, you can register a new zone with create_domain or list existing domains using list_domains.

You then pinpoint the exact record using create_domain_record, and if you're just checking what you own, list_domains does the trick. Finally, when your setup changes, set up an automated alert endpoint by running create_webhook, and test that connection with test_webhook.

Storage and Data Management
Don't sweat storage. You can allocate a new block volume using create_volume, or check what volumes are already waiting for you with the system’s list functions. When an instance needs data, connect a stored volume using attach_volume; when it's time to clean house, use detach_volume.

Billing and Resource Oversight
Knowing what you spend is key. You can get a detailed report of all hourly charges for the current billing cycle by calling get_charges. Want to know your limits? Use get_quota to check resource consumption against set quotas. For a full picture, run list_regions to see where you can deploy resources, and list_networks shows every private network segment you’ve established.

Team and User Management
Managing who has access is just as important. You can create new organizational groups with create_team, then add specific users to that team using add_team_member. If someone's status needs updating, change it with update_team_member_status.