Tailscale Compute runs on your own machine — the client starts it, so there is no endpoint to ping. 95 installs a week from npm. Last commit 15 Aug 2026.
Run builds, tests, and benchmarks on your own Tailscale compute node.
We read the source, 3 d ago · rules 995398eba69f
What this server is able to do. For an MCP server this is often the job itself — a terminal server runs commands because that is what it is for. Listed so you know what you are plugging in, not as an accusation.
child = spawn(request.executable, [...request.arguments], {
Is this your server and something here is wrong? Tell us — corrections are free and do not require a plan.
We found places where it runs commands, builds paths or queries from values it is given. None of that is a flaw by itself — it becomes one when the code changes, and code changes quietly between releases. We re-read it on every one.
This server runs on your own machine — install it with the package manager and the client starts it for you. Package name taken from the official registry entry.
claude mcp add tailscale-compute -e 'TAILSCALE_COMPUTE_HOST=<TAILSCALE_COMPUTE_HOST>' -- npx -y @dylantirandaz/tailscale-compute-mcp
{
"mcpServers": {
"tailscale-compute": {
"command": "npx",
"args": [
"-y",
"@dylantirandaz/tailscale-compute-mcp"
],
"env": {
"TAILSCALE_COMPUTE_HOST": "<TAILSCALE_COMPUTE_HOST>"
}
}
}
}
[mcp_servers.tailscale-compute]
command = "npx"
args = ["-y", "@dylantirandaz/tailscale-compute-mcp"]
env = { TAILSCALE_COMPUTE_HOST = "<TAILSCALE_COMPUTE_HOST>" }
{
"mcpServers": {
"tailscale-compute": {
"command": "npx",
"args": [
"-y",
"@dylantirandaz/tailscale-compute-mcp"
],
"env": {
"TAILSCALE_COMPUTE_HOST": "<TAILSCALE_COMPUTE_HOST>"
}
}
}
}
{
"servers": {
"tailscale-compute": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@dylantirandaz/tailscale-compute-mcp"
],
"env": {
"TAILSCALE_COMPUTE_HOST": "<TAILSCALE_COMPUTE_HOST>"
}
}
}
}
Put your own value in place of
<TAILSCALE_COMPUTE_HOST> (SSH user and direct Tailscale IP address or full MagicDNS name).
The author’s registry entry marks it as required.
Optional settings from the registry entry, left out of the block above:
TAILSCALE_COMPUTE_LOCAL_ROOT (Absolute local project path. The default is the MCP process directory), TAILSCALE_COMPUTE_REMOTE_ROOT (Managed workspace root on the remote node. The default is .cache/tailscale-compute-mcp), TAILSCALE_COMPUTE_REMOTE_SHELL (Remote shell: auto, /bin/sh, /bin/bash, or /bin/zsh), TAILSCALE_COMPUTE_CONNECT_TIMEOUT_SECONDS (SSH connection timeout from 1 through 60 seconds), TAILSCALE_COMPUTE_AUDIT_LOG (Local audit log path. The default is ~/.config/tailscale-compute-mcp/compute-audit.log).
Run GPU acceptance tests on your NVIDIA GPU nodes over SSH: DCGM, GEMM, HBM/PCIe bandwidth, NCCL.
Create, run and analyze A/B tests and personalizations on your site with Mida.
Create and run A/B tests, patches, split tests and feature flags on your website
A real Mac for your agent: build, test, run, and tap through iOS apps on the simulator.
Test and compare prompts across any AI provider. Bring your own keys.
Runs Foundry tests and builds; returns compact, structured failures to AI coding agents
Your AI dev team: agents that build, test, review and ship, each on its own cloud machine.
A/B testing MCP server for pagent: create, run and review A/B tests on your live website.
Answers built from our own checks of this server.