VerifyAX MCP server

This is a helper that connects your AI assistant to VerifyAX, a platform for testing how well AI agents behave. Once it is set up, you can just describe what you want in plain words and your assistant does the clicking for you. It is handy for anyone who checks or compares AI agents without wanting to write code.

What this MCP server does

You tell your AI what you want, like registering an agent or running an evaluation. The AI uses this helper, which talks to the VerifyAX platform using your account. For longer jobs, like generating a scenario or running an evaluation, the helper waits until the work finishes and then brings back the result. You see the answer right in your chat, such as scores, a list of agents, or a cost estimate. You never have to write any code or open a separate dashboard.

Click to zoom

What you can do with it

• Register a new agent and check that it is reachable
• List all the agents in your workspace
• Generate a test scenario from skill tags
• Run an evaluation of an agent against a scenario
• Read scores and transcripts from a past run
• Preview how many credits a run will cost
• Delete an agent or scenario you no longer need

Try asking your AI

• “Register my agent at https://my-agent.example.com and confirm it is reachable”
• “Generate an info_exchange scenario tagged empathy and coordination”
• “Evaluate agent X against scenario Y and summarize the scores”
• “How many credits will it cost to run this scenario against my agent”

What it gives back to you

You get answers written in the chat, like a list of your agents, a summary of evaluation scores, or a short transcript from a run. It can also show credit usage and cost estimates as numbers. When it creates or deletes something, it tells you what it did. Nothing appears in a separate file unless you ask your AI to save it.

Before you start

Who it's for

People who test or compare AI agents, like QA testers, product managers, and support teams, who want to work in plain language instead of code.