wtf is securevector-eu-demo?

This project is a test script for people already using SecureVector, a security monitoring tool for AI agents. Rather than something end users interact with directly, it is a companion setup that automatically deploys a copy of the SecureVector engine to cloud infrastructure in Europe, checks that it works correctly, and then tears everything back down again. It uses Terraform, a tool for describing and creating cloud infrastructure through code, to spin up the necessary pieces on Amazon Web Services: a running container, a load balancer to route traffic to it, and related resources, all placed in an EU region such as Ireland or Frankfurt. Once deployed, the script checks that the health endpoint responds, that the main analysis endpoint correctly rejects requests without a valid access token and accepts them with one, and that the engine reports its own configuration correctly. It can optionally also run a real example agent built with LangChain, a framework for building AI applications, against the freshly deployed engine to prove it works end to end in practice. A specific goal of the setup is data residency: keeping prompt and output data physically within the EU rather than letting it leave the region, which matters for organizations that need to meet European data handling requirements. The readme is direct about the risks of running this outside of its intended test use though, warning that by default it creates an internet facing endpoint using plain, unencrypted web traffic, and that skipping the access token setting would leave the endpoint completely open to anyone. The provided script always sets a strong random token for each run and cleans up its resources automatically afterward. Using it requires Terraform, AWS account credentials, and a few common command line tools already installed, and running it will incur small, short lived AWS costs while the infrastructure is briefly running.