Toolroute runs on your own machine — the client starts it, so there is no endpoint to ping. 3 177 installs a week from npm. Last commit 2 Jul 2026.
Route AI agent tasks to the best MCP server and LLM, scored on 132+ real benchmark executions.
Today is the operative word: we check Toolroute 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 toolroute -- npx -y @toolroute/sdk
{
"mcpServers": {
"toolroute": {
"command": "npx",
"args": [
"-y",
"@toolroute/sdk"
]
}
}
}
[mcp_servers.toolroute]
command = "npx"
args = ["-y", "@toolroute/sdk"]
{
"mcpServers": {
"toolroute": {
"command": "npx",
"args": [
"-y",
"@toolroute/sdk"
]
}
}
}
{
"servers": {
"toolroute": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@toolroute/sdk"
]
}
}
}
Read-only MCP server for the OPERANT AI operating-agent calibration benchmark.
LoopPause MCP server — pause agent execution and route human approval requests
Find the best eligible service for an agent task through one task-first MCP tool.
Read-only lookup of tested agent tools and MCP servers: scores, test evidence, deprecation.
Evaluate an MCP server with scout and verify scout attestations, from inside the agent. Read-only.
Read-only MCP server over 50 book-based agent skills: routes a task to the right framework.
Pay-per-call AI evaluation MCP server. Score LLM outputs against benchmark rubrics via Workers AI.
MCP server for AI agents to report infrastructure needs they encounter during task execution
Answers built from our own checks of this server.