MCP server · Security

Agent-Aegis governance MCP server

Adds a safety layer to your AI's tool calls, with rules, approvals and an audit log.

Security For it and development Security

Agent-Aegis is a helper that sits between your AI and the tools it uses. It checks each tool call against rules you write, can pause and ask for approval, and keeps a record of what happened. It is handy for anyone letting an AI agent touch real systems like a CRM, a database or a website.

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 an app or service. This one connects your AI to Agent-Aegis, a governance layer that watches over the tool calls your AI makes. So when your AI wants to do something, the request goes through Aegis first, which checks the rules and either lets it through, asks you to approve, or blocks it.

What this MCP server does

You ask your AI to do something, like update a customer record. Your AI decides which tool to use and makes the call. That call passes through Agent-Aegis, which looks at your rules written in a simple YAML file. Aegis scores the risk, checks for things like leaked personal data or sneaky instructions, and decides: allow it, ask you to approve, or block it. Then it writes the whole thing to a tamper-evident log so you can see what happened later.

Flow diagram: you ask your AI “Update the CRM record for customer 4821”, the Agent-Aegis governance MCP server keeps repeating: your AI asks to act, aegis checks the rules, aegis decides, aegis writes the log, and you get back allowed, blocked, or waiting.

Click to zoom

What you can do with it

Block prompt injection attempts before they reach your AI
Warn you when personal data like emails or card numbers show up
Require your approval before risky actions like bulk updates or deletes
Keep a signed, tamper-evident log of every tool call
Set cost budgets and rate limits so an agent cannot run wild
Scan your code to find AI calls that have no governance at all
Track which agent handed a task to which other agent

Try asking your AI

“Update the CRM record for customer 4821 with the new phone number”
“Delete all inactive accounts from last year”
“Send a summary email to the whole sales team”
“Run the monthly report and post the numbers to Slack”

What it gives back to you

You get back a clear answer in the chat: whether the action was allowed, blocked, or is waiting for your approval. If it was blocked, Aegis tells you why, such as a risk score or a rule that matched. You can also ask for the audit log and get a table of recent actions with their risk level and decision. If personal data was found, you see a warning instead of the raw values.

Before you start

What you need

Python 3.10 or newer installed on your computer
The agent-aegis package installed (pip install agent-aegis)
A policy file written in YAML that says what is allowed, approved, or blocked

Good to know

Aegis can block or hold back actions your AI tries to take, and its audit log records what your agents do, so treat the policy file and log as sensitive.

Install it with your AI

Add Agent-Aegis governance 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 Agent-Aegis governance 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.