tirekicker
Briefing 01 · October 2026

We tested every remote MCP server in the registry. One in ten is dead.

24,258 servers, each one connected to and asked what it offers. The first Tirekicker briefing says how many work, how many hide a login, and how quickly a listing goes stale. It is free to people on the waiting list.

Join the waiting list

Free. The briefing is sent by email.

45%

of servers listed more than six months ago
no longer work. For new listings it is 3%.

From the briefing

A listing tells you a server exists. It doesn't tell you it still works.

We connected to every remote server in the official registry on 5 October 2026 and recorded what happened. A few of the things we found:

  • 72%of the servers that demanded credentials say nothing about it in their listing.
  • 1,565servers that anyone can connect to expose a tool their own author labels destructive.
  • 21%of all remote listings come from just three publishers.
  • 64%of the 300 most-downloaded local servers start cleanly on a fresh machine.

The briefing has the full figures, eight findings, and a checklist of what to require before letting an agent depend on an MCP server.

Method

What we do to every server

The same procedure for every server, run by software, with the raw output kept. No judgement calls and no favourites.

  1. Install it cleanA throwaway container with no credentials, capped memory and a hard time limit. Nothing carries over between servers.
  2. Start itWe record whether it starts, how long it takes, and what it prints if it doesn't.
  3. Ask what it offersWe complete the protocol handshake and request its tool list, then fingerprint that list.
  4. Do it againOne run is a snapshot. Re-running is what shows what broke, what recovered, and whose tools quietly changed. Two runs so far.
What this does not tell you
  • We do not call any tool yet, so a server that starts may still have tools that fail.
  • A server that needs an API key is reported as "needs key", not as broken. We test without credentials.
  • This is not a security audit. Use a scanner for that, alongside this.
Independence

Nobody pays for a result

No paid rankingsA server's result is whatever the test recorded. It cannot be bought, sponsored or improved by asking.
Open method, raw outputThe procedure and the recorded output are published with every result, so you can check the work instead of trusting us.
Corrections in publicIf a result is wrong, we re-test and say so. Authors can request a re-test; the new result is published whatever it shows.