DigitalOcean MCP
Manage Droplets, apps, databases, networking and more on DigitalOcean.
Labels are provisional (from the project README/docs, checked Oct 27, 2026). Not yet tested in our sandbox.
Read-only switch
No global read-only switch documented; services can be selected. Scope via a custom-scoped API token.
Preset
Claude Code
{
"mcpServers": {
"digitalocean": {
"command": "npx",
"args": ["@digitalocean/mcp", "--services", "apps"],
"env": {
"DIGITALOCEAN_API_TOKEN": "${DIGITALOCEAN_API_TOKEN}"
}
}
}
}{
"permissions": {
"ask": ["mcp__digitalocean"]
}
}ask makes Claude Code prompt before every tool from this server.
Codex
# Generated by redpresets.dev on 2026-11-03. Unverified preset. Also use a read-only credential. Not a security guarantee.
[mcp_servers.digitalocean]
command = "npx"
args = ["@digitalocean/mcp", "--services", "apps"]
env_vars = ["DIGITALOCEAN_API_TOKEN"]
default_tools_approval_mode = "prompt"default_tools_approval_mode = "prompt" asks before every tool call.
Cursor
{
"mcpServers": {
"digitalocean": {
"type": "stdio",
"command": "npx",
"args": ["@digitalocean/mcp", "--services", "apps"],
"env": {
"DIGITALOCEAN_API_TOKEN": "${env:DIGITALOCEAN_API_TOKEN}"
}
}
}
}Don't add write tools to mcpAllowlist in .cursor/permissions.json. Leave them on approval. Cursor's docs say the allowlist is not a security boundary.
Credential layer
Create a DigitalOcean custom-scoped token with only read scopes.
Facts
- Vendor
- DigitalOcean official
- License
- MIT
- GitHub stars
- 138
- Latest release
- v1.6.1 (Oct 23, 2026)
- Last push
- Oct 27, 2026
- Status
- Active
- Last checked
- Oct 27, 2026
Sources
- README · checked Oct 27, 2026
Spotted something wrong? Labels change when vendors change. Suggest a correction
Read-only alternatives in Cloud
Metrics, alarms and Logs Insights queries from CloudWatch.
EKS cluster management, Kubernetes resources and troubleshooting.
Runs AWS CLI commands for the agent across services.