Tailscale RMCP MCP server
Tailscale device, route, DNS, key, user, and ACL management over MCP and CLI.
5 stars46 downloads/wk
Reviews
Write oneNobody has reviewed Tailscale RMCP yet.
If you have run it, two minutes of your experience saves the next person an afternoon.
Tailscale RMCP tools
Tool list not cached yet. `describe` through the gateway fetches it live.
Public scan report
scanner v0.1.9 · 2026-09-20 · same rubric, same numbers if you re-run it
- Code scan3 source files scanned15/25
- Live reliabilityremote reachable in 268ms (auth required)20/20
- –Tool poisoningtools not inspected (endpoint requires auth); not countedn/a
- Auth qualityauth enforced (unknown)9/15
- Maintenancelast push 26 days ago15/15
- Maintainer identitynamespace and repository owner differ; website matches verified namespace6/10
Findings (2)
- mediumnpm install lifecycle script present
install.script - mediumnpm install lifecycle script present
install.scriptpackage.json: …ck": "node scripts/sync-readme.js", "postinstall": "node scripts/install.js", "test": …
What the publisher says
From the Tailscale RMCP repository's README, as published. We do not edit it. Read it on GitHub
tailscale-rmcp
Tailscale device, route, DNS, key, user, and ACL management over MCP and CLI.
It exposes one MCP tool, tailscale, plus the rtailscale CLI. Agents can list devices, inspect routes, read API keys, ACL policy, DNS settings, and users, authorize devices, and delete devices when the destructive gate is explicitly enabled.
30-second path: set TAILSCALEAPIKEY, then run npx -y @dinglebear/rtailscale devices --json -> start loopback HTTP with TAILSCALEMCPHOST=127.0.0.1 npx -y @dinglebear/rtailscale serve -> call tools/call with {"action":"devices"}.
Status: operational RMCP upstream-client server. Write-capable for device authorization; destructive device deletion requires both server opt-in and caller confirmation. HTTP MCP supports loopback dev mode, static bearer tokens, and Google OAuth through lab-auth.
Not for: replacing the Tailscale admin console, bypassing Tailscale account permissions, operating multiple unrelated tailnets from one trust boundary, storing API keys for callers, arbitrary WireGuard control, or passing Tailscale API keys through MCP tool arguments.
Contents
- Naming
- Capabilities And Boundaries
- Install
- Quickstart
- Client Configuration
- Runtime Surfaces
- MCP Tool Reference
- CLI Reference
- Configuration
- Authentication
- Safety And Trust Model
- Architecture
Naming
The repo, crate, and npm package use the RMCP family name. The shipped binary is rtailscale to avoid shadowing the official tailscale CLI.
Capabilities And Boundaries
ID.
- List devices and inspect a single device by node ID or legacy numeric device
users.
- Read subnet routes, API keys, ACL policy, DNS/MagicDNS settings, and tailnet
also passes explicit confirmation.
- Authorize a device for the tailnet.
- Delete a device only when TAILSCALEALLOWDESTRUCTIVE=true and the caller
- Provide setup and doctor commands for local plugin/runtime checks.
Install
Shortened. The full README is on GitHub.
Nothing above is checked by us. What we check is on the safety report.
Install directly
claude mcp add --transport http rtailscale https://ts.dinglebear.ai/mcp
Tailscale RMCP: common questions
- Is Tailscale RMCP MCP server safe?
- Mostly: it is graded B (76/100). Read the Tailscale RMCP safety report
- How do I install Tailscale RMCP?
- It runs remotely at ts.dinglebear.ai. Add it to Claude Code, Claude Desktop or Cursor with the snippets above, or call it through the mcp.market gateway without installing anything.
- Does Tailscale RMCP need an API key?
- Yes. The registry entry asks for
TAILSCALE_API_KEY,Authorization. - Is Tailscale RMCP maintained?
- The last commit was 26 days ago (2026-08-25). The latest release is v0.1.4.
- Is Tailscale RMCP up?
- 0% of our last 6 checks got an answer. We check remote servers about four times a day.