mcpField
V
vultr

vultr

github.com/evolv3-aiOther140 installsUpdated 14w ago
Repository

About vultr

A deployment skill for provisioning and managing Vultr servers with vultr-cli and SSH. It guides setup from collecting configuration through instance verification, and includes Kubernetes autoscaling steps.

Key features

  1. 01Collects required deployment parameters
  2. 02Selects server profiles by purpose
  3. 03Creates instances and firewall groups
  4. 04Verifies SSH access
  5. 05Documents VKE node autoscaling

Use cases

  1. 01Provisioning a Vultr server for Coolify or Kasm
  2. 02Choosing a Vultr plan for a workload
  3. 03Creating a Kubernetes cluster with autoscaling

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/evolv3-ai/vibe-skills --skill vultr

After installing, type/vultrin Claude Code to confirm it loaded.

Bundled filesassets/env-templatereferences/OPERATIONS.md

Details

From the project README

Vultr Infrastructure

Status: Production Ready | Dependencies: vultr-cli, SSH key pair


Navigation

  • Operations, troubleshooting, config, and cost snapshot: references/OPERATIONS.md

Step 0: Gather Required Information (MANDATORY)

STOP. Before ANY deployment commands, collect ALL parameters from the user.

Copy this checklist and confirm each item:

Required Parameters:
- [ ] SERVER_NAME      - Unique name for this server
- [ ] VULTR_REGION     - Region (ewr, ord, dfw, lax, lhr, fra, sgp, etc.)
- [ ] VULTR_PLAN       - Plan ID (see profiles below)
- [ ] SSH_KEY_NAME     - Name of SSH key in Vultr
- [ ] SSH_KEY_PATH     - Path to local SSH private key (default: ~/.ssh/id_rsa)

Deployment Purpose (determines recommended profile):
- [ ] Purpose: coolify / kasm / both / custom
      coolify → vc2-2c-4gb ($24/mo)
      kasm    → vc2-4c-8gb ($48/mo)
      both    → vc2-8c-32gb ($192/mo)
      custom  → Ask for specific plan

Recommended profiles by purpose:

PurposePlanvCPURAMMonthly
coolifyvc2-2c-4gb24GB$24
kasmvc2-4c-8gb48GB$48
bothvc2-8c-32gb832GB$192

DO NOT proceed to Prerequisites until ALL parameters are confirmed.


Prerequisites

Before using this skill, verify the following:

1. Vultr CLI Installed
vultr-cli version

If missing, install with:

# macOS
brew install vultr/vultr-cli/vultr-cli

# Linux (download binary)
curl -sL https://github.com/vultr/vultr-cli/releases/latest/download/vultr-cli_linux_amd64.tar.gz | tar xz
sudo mv vultr-cli /usr/local/bin/

# Windows (scoop)
scoop bucket add vultr https://github.com/vultr/scoop-vultr.git
scoop install vultr-cli

# Go install
go install github.com/vultr/vultr-cli/v3@latest
2. Vultr Account & API Key

If you don't have a Vultr account:

Sign up at: https://www.vultr.com/?ref=YOUR_REFERRAL_CODE

Disclosure: This is a referral link. You'll receive $100 in credit, and the skill author receives $35 after you spend $25. Using this link helps support the development of these skills.

Get API Key: https://my.vultr.com/settings/#settingsapi

Create a Personal Access Token with All permissions.

3. vultr-cli Configured
vultr-cli account

If it shows an error, configure with:

# Set API key
vultr-cli config set api-key YOUR_API_KEY_HERE

# Or via environment variable
export VULTR_API_KEY="your_api_key_here"
vultr-cli account
4. SSH Key Pair
ls ~/.ssh/id_rsa.pub

If missing, generate with:

ssh-keygen -t rsa -b 4096 -f ~/.ssh/id_rsa -N ""
5. SSH Key Uploaded to Vultr
vultr-cli ssh-key list

If empty, upload with:

vultr-cli ssh-key create --name "my-key" --key "$(cat ~/.ssh/id_rsa.pub)"
6. Test Authentication
vultr-cli regions list

If this fails: API key may be invalid. Create a new one.


Server Profiles

Coolify/Kasm Deployments
ProfilePlanvCPURAMDiskMonthly Cost
coolifyvc2-2c-4gb24GB80GB$24
kasmvc2-4c-8gb48GB160GB$48
bothvc2-8c-32gb832GB640GB$192
High-Frequency (NVMe, Best Performance)
ProfilePlanvCPURAMDiskMonthly Cost
hf-smallvhf-2c-4gb24GB64GB NVMe$24
hf-mediumvhf-4c-16gb416GB256GB NVMe$72
hf-largevhf-6c-24gb624GB384GB NVMe$108
AMD High Performance (Dedicated vCPU)
PlanvCPURAMDiskMonthly Cost
vhp-2c-4gb-amd24GB60GB NVMe$30
vhp-4c-8gb-amd48GB120GB NVMe$60
vhp-8c-16gb-amd816GB240GB NVMe$120

Deployment Steps

Step 1: Set Environment Variables
export VULTR_REGION="ewr"                  # See regions below
export VULTR_PLAN="vc2-2c-4gb"             # See profiles above
export VULTR_OS_ID="1743"                  # Ubuntu 22.04 x64
export SERVER_NAME="my-server"
export SSH_KEY_NAME="my-key"
Region options
CodeLocationRegion
ewrNew JerseyUS East
ordChicagoUS Central
dfwDallasUS South
laxLos AngelesUS West
seaSeattleUS Northwest
atlAtlantaUS Southeast
miaMiamiUS Southeast
ytoTorontoCanada
lhrLondonUK
amsAmsterdamNetherlands
fraFrankfurtGermany
cdgParisFrance
nrtTokyoJapan
sgpSingaporeAsia
sydSydneyAustralia
blrBangaloreIndia

Run vultr-cli regions list for full list.

OS ID reference
OS IDOperating System
1743Ubuntu 22.04 x64
2136Ubuntu 24.04 x64
477Debian 12 x64
447Debian 11 x64
215CentOS 7 x64

Run vultr-cli os list for full list.

Step 2: Get SSH Key ID
SSH_KEY_ID=$(vultr-cli ssh-key list | grep "$SSH_KEY_NAME" | awk '{print $1}')
echo "SSH Key ID: $SSH_KEY_ID"

# Verify
if [ -z "$SSH_KEY_ID" ]; then
  echo "ERROR: SSH key '$SSH_KEY_NAME' not found. Upload it first."
  exit 1
fi
Step 3: Create Firewall Group
# Create firewall group
FW_GROUP_ID=$(vultr-cli firewall group create --description "my-firewall" | grep -oP 'ID: \K\S+')
echo "Firewall Group ID: $FW_GROUP_ID"

# Allow SSH
vultr-cli firewall rule create --id "$FW_GROUP_ID" --protocol tcp --port 22 --subnet 0.0.0.0/0

# Allow HTTP/HTTPS
vultr-cli firewall rule create --id "$FW_GROUP_ID" --protocol tcp --port 80 --subnet 0.0.0.0/0
vultr-cli firewall rule create --id "$FW_GROUP_ID" --protocol tcp --port 443 --subnet 0.0.0.0/0

# Allow Coolify ports
vultr-cli firewall rule create --id "$FW_GROUP_ID" --protocol tcp --port 8000 --subnet 0.0.0.0/0
vultr-cli firewall rule create --id "$FW_GROUP_ID" --protocol tcp --port "6001:6002" --subnet 0.0.0.0/0

# Allow KASM ports
vultr-cli firewall rule create --id "$FW_GROUP_ID" --protocol tcp --port 8443 --subnet 0.0.0.0/0
vultr-cli firewall rule create --id "$FW_GROUP_ID" --protocol tcp --port 3389 --subnet 0.0.0.0/0
vultr-cli firewall rule create --id "$FW_GROUP_ID" --protocol tcp --port "3000:4000" --subnet 0.0.0.0/0
Step 4: Create Instance
vultr-cli instance create \
  --region "$VULTR_REGION" \
  --plan "$VULTR_PLAN" \
  --os "$VULTR_OS_ID" \
  --ssh-keys "$SSH_KEY_ID" \
  --firewall-group "$FW_GROUP_ID" \
  --label "$SERVER_NAME" \
  --hostname "$SERVER_NAME"
Step 5: Wait and Get Instance IP
# Wait for instance to be ready
echo "Waiting for instance to be active..."
sleep 30

# Get instance ID
INSTANCE_ID=$(vultr-cli instance list | grep "$SERVER_NAME" | awk '{print $1}')
echo "Instance ID: $INSTANCE_ID"

# Get IP address
SERVER_IP=$(vultr-cli instance get "$INSTANCE_ID" | grep "Main IP" | awk '{print $3}')
echo "SERVER_IP=$SERVER_IP"
Step 6: Wait for Server Ready
# Wait for SSH to be available (typically 60-90 seconds)
echo "Waiting for server to be ready..."
until ssh -o ConnectTimeout=5 -o StrictHostKeyChecking=no root@$SERVER_IP "echo connected" 2>/dev/null; do
  sleep 5
done
echo "Server is ready!"
Step 7: Verify Connection
ssh root@$SERVER_IP "uname -a && free -h && df -h /"
Step 8: Output for Downstream Skills
# Vultr only offers x86 architecture
SERVER_ARCH="amd64"

# Save to .env.local for downstream skills
echo "SERVER_IP=$SERVER_IP" >> .env.local
echo "SSH_USER=root" >> .env.local
echo "SSH_KEY_PATH=~/.ssh/id_rsa" >> .env.local
echo "SERVER_ARCH=$SERVER_ARCH" >> .env.local
echo "COOLIFY_SERVER_IP=$SERVER_IP" >> .env.local
echo "KASM_SERVER_IP=$SERVER_IP" >> .env.local

echo ""
echo "Instance deployed successfully!"
echo "  IP: $SERVER_IP"
echo "  Arch: $SERVER_ARCH"
echo "  SSH: ssh root@$SERVER_IP"

Verify Deployment

ssh root@$SERVER_IP "echo 'Vultr instance connected successfully'"

Kubernetes Auto-Scaling

Vultr Kubernetes Engine (VKE) supports node pool autoscaling.

Create VKE Cluster with Autoscaling
# Create cluster
vultr-cli kubernetes create \
  --region "$VULTR_REGION" \
  --version "v1.28.2+1" \
  --label "my-k8s"

# Get cluster ID
CLUSTER_ID=$(vultr-cli kubernetes list | grep "my-k8s" | awk '{print $1}')

# Create autoscaling node pool
vultr-cli kubernetes node-pool create "$CLUSTER_ID" \
  --plan "vc2-2c-4gb" \
  --label "worker-pool" \
  --quantity 2 \
  --min-nodes 1 \
  --max-nodes 5 \
  --auto-scaler true
How It Works
  • Vultr automatically adds nodes when workload increases
  • Nodes are removed when demand decreases
  • Specify min-nodes and max-nodes to control scaling range

Cleanup

Warning: This is destructive and cannot be undone.

# Delete instance
INSTANCE_ID=$(vultr-cli instance list | grep "$SERVER_NAME" | awk '{print $1}')
vultr-cli instance delete "$INSTANCE_ID"

# Delete firewall group
FW_GROUP_ID=$(vultr-cli firewall group list | grep "my-firewall" | awk '{print $1}')
vultr-cli firewall group delete "$FW_GROUP_ID"

# Optionally delete SSH key
# vultr-cli ssh-key delete "$SSH_KEY_ID"

Operations

Troubleshooting, best practices, configuration variables, and cost snapshots are in references/OPERATIONS.md.


Logging Integration

When performing infrastructure operations, log to the centralized system:

# After provisioning
log_admin "SUCCESS" "operation" "Provisioned Vultr instance" "id=$INSTANCE_ID provider=Vultr"

# After destroying
log_admin "SUCCESS" "operation" "Deleted Vultr instance" "id=$INSTANCE_ID"

# On error
log_admin "ERROR" "operation" "Vultr deployment failed" "error=$ERROR_MSG"

See admin skill's references/logging.md for full logging documentation.


References

Comments

Markdown supported