Hyperlight runs on your own machine — the client starts it, so there is no endpoint to ping. 170 installs a week from pypi. Last commit 6 Nov 2026.
Runs Python in a local, hardware-isolated micro-VM that keeps its state between calls
We read the source, 23 h ago · rules 995398eba69f
Found in continuous integration, deployment or infrastructure files, or in a neighbouring package of the same monorepo. None of this is installed when you add the server: it describes how the project is built and released. We list it because a leaked key in a build pipeline is still a real problem, but it is not something this server does on your machine.
cp = subprocess.run(
exec(compile(ast.Module(body=[node], type_ignores=[]), "<import>", "exec"), {})
control = os.path.join(temporary, _CONTROL_PREFIX + secrets.token_hex(8))
Is this your server and something here is wrong? Tell us — corrections are free and do not require a plan.
Today is the operative word: we check Hyperlight every 15 minutes and re-read its code on every release. Watch it and you find out the day that stops being true.
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 mcp-server-hyperlight -- uvx mcp-server-hyperlight
{
"mcpServers": {
"mcp-server-hyperlight": {
"command": "uvx",
"args": [
"mcp-server-hyperlight"
]
}
}
}
[mcp_servers.mcp-server-hyperlight]
command = "uvx"
args = ["mcp-server-hyperlight"]
{
"mcpServers": {
"mcp-server-hyperlight": {
"command": "uvx",
"args": [
"mcp-server-hyperlight"
]
}
}
}
{
"servers": {
"mcp-server-hyperlight": {
"type": "stdio",
"command": "uvx",
"args": [
"mcp-server-hyperlight"
]
}
}
}
Single exec tool running caller Python in a network-isolated locked-down sandbox.
Ephemeral MicroVM-isolated code execution for AI agents. Fresh VM per call, hard-purged after.
Find open-weight AI models that fit your hardware, with weight sizes, memory and sources.
AI assistant access to PicoCalc MicroPython hardware over USB.
Install and run open-source OSINT tools in a local sandbox, asking before runs that touch a target.
Live UK, Germany, Canada and US prices for GPUs and Macs that run open AI models. No sign-up.
Check if a task runs locally vs cloud. Save money on calls that don't need cloud inference.
Local code graph for TypeScript, Python and Swift: who calls this, what breaks if I change it.
Answers built from our own checks of this server.