MCP server · Security

Aggrete MCP server

Your AI checks each request against your company's code of conduct and refuses the ones that are not allowed.

Aggrete sits between your AI and the tools it uses, like Slack, Drive and GitHub. Before your AI fetches anything, Aggrete checks the request against a written set of rules and says yes or no. It is handy for teams that worry about the AI pulling together information it should not.

What is an MCP server? The 30-second version

On its own, your AI can only chat with you. An MCP server is a small helper program that gives your AI a new skill or a connection to another service. This one connects your AI to your work apps through a set of rules, so the AI can look things up or make changes there when you ask. Aggrete adds one more thing on top: it checks each request against your code of conduct before it happens.

What this MCP server does

You ask your AI for something, like a list of recent hires or a budget breakdown. Your AI sends that request through Aggrete instead of going straight to Slack, Drive or GitHub. Aggrete reads your rules file and looks at what this person has already pulled today, then decides if the request is allowed. If it is fine, Aggrete fetches the data and passes it back. If the request would combine information in a way your rules forbid, Aggrete refuses it before anything is fetched.

Flow diagram: you ask your AI “Show me the new hires this month”, the Aggrete MCP server connects it to Aggrete, and you get back allowed or refused, with the reason.

Click to zoom

What you can do with it

Refuse a request before any data is fetched, based on your written rules
Remember what each person has already pulled today and flag risky combinations
Hide tools from people who are not allowed to use them
Blur emails, card numbers and tokens out of results before your AI sees them
Block any write action after a session has read untrusted content
Pause a request and ask a human to approve it before it goes through
Keep a tamper-evident log of every decision for later review

Try asking your AI

“Check this plan before running it: pull recent joiners, budget roles and the on-call draft”
“List the scenarios I can try on this proxy”
“Show me the audit log for today”
“Ask the HR clause owner to approve this request”

What it gives back to you

You get a clear verdict in the chat: allowed or refused, with the rule that was triggered. If allowed, you also get the data itself, with sensitive bits already blurred out. If refused, you get the reason and the rule code, so you know why. For held requests, you get a note that it is waiting for a human to approve.

Before you start

What you need

Python installed on your computer
A rules file written in YAML that describes your code of conduct
A config file that lists which apps to connect
For real use: an HTTP setup with OAuth so the proxy is the only path to your apps

Good to know

It can refuse or hold your requests, and once a request is refused the data is not fetched, so plan your rules carefully before turning it on for real.

Install it with your AI

Add Aggrete 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 Aggrete 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.