Complete AI Training

MCP server · Security

Rails MCP server

by jaimenbell

Gives your AI a safety rail: it can check whether an action is risky and keep a record of spending and human approvals.

Flow diagram: you ask your AI “Is deploy_prod a gated action?”, on your own computer the Rails MCP server works with your own computer, and you get back A short answer in the chat.

Rails MCP is a small helper you run on your own computer. It gives your AI a way to check whether an action should be treated as risky, keep a list of spending intentions, and record when a human has approved something. It is handy for people who let an AI assistant do real work and want a paper trail, not just a chat.

What is an MCP server? The 30-second version

On its own, your AI can only chat. An MCP server is a small helper program that gives your AI a new skill or a connection to a service. This one connects your AI to a safety registry and two record books that live on your own machine. When you ask your AI something, it can use this helper to look up whether an action is risky, or to write down a spending intention.

What this MCP server does

You ask your AI a question, like whether a certain action is safe or how much has been spent this month. The AI uses this helper to check its registry or its ledgers. The helper reads your config file and your record files, and returns a verdict, a number, or a short summary. Nothing here blocks anything by itself. It gives you the information and the record, and you decide what to do with it.

Flow diagram: you ask your AI “Is deploy_prod a gated action?”, on your own computer the Rails MCP server works with your own computer, and you get back A short answer in the chat. Click to zoom

What you can do with it

  • Check whether an action type is treated as risky (it always says GATED)
  • See whether an action type is listed in your registry
  • Get a short fingerprint of your current registry
  • See who last signed off on your registry, and when
  • Write down a spending intention with amount, vendor, and purpose
  • Check total spending in the last 30 days against a limit

Try asking your AI

  • “Is deploy_prod a gated action?”
  • “Is deploy_prod registered in my rails config?”
  • “Record a spend intention of 40 dollars for OpenAI, purpose: embeddings batch”
  • “Have we spent more than 200 dollars in the last 30 days?”

What it gives back to you

You get short answers in the chat. A verdict like GATED, a yes or no about whether something is registered, a 12-character fingerprint, a sign-off record or null, or a spending total compared to your limit. When you record a spend intention, you get a confirmation that the line was appended.

Before you start

What you need

  • Python 3.10 or newer
  • The rails-mcp package installed (pip install rails-mcp)
  • A config file listing your action types (optional; without it the registry is empty)
  • A place on disk for the two ledger files (the default is ./rails_data/)

Good to know

This does not block or stop anything by itself, and the sign-off record is not tamper-proof: anyone with access to the ledger file can add a fake line.

Install it with your AI

Add Rails MCP server to your AI, no technical skills needed

You don't install anything by hand. You copy one prompt, paste it into an AI that can work on your computer, and it checks, installs and connects the server for you, asking you when it needs something.

Sign in to get the install prompt

Members get a ready-made prompt that lets the Claude desktop app check Rails MCP server, install it and connect it for them, step by step. You don't need any technical skills: you copy, paste and answer a few questions. Your connected AI can also find and install any of the 4,066 MCP servers here for you.

Sign in Become a member

Who it's for

Teams and individuals who let AI agents take real actions and want a simple, self-hosted record of what is risky, what was spent, and who approved what.