Hi all! for the past few months I've been building simulated sandboxes as an approach to buid safer

Hi all! for the past few months I've been building simulated sandboxes as an approach to buid safer ai agents. If you want to see how your ai agents will behave when interacting with live APIs, databases or MCP tools. One bad call can burn through rate limits or accidentally wipe an entire database. To solve this we're building Digital Twins which are simulated sandbox environments of different APIs. You can use these to test your agents without touching live endpoints or having to worry about maintain your testing infrastructure. Even with Jev, you still need a testing framework to build production agents and pome can help with this 😁 GitHub: https://github.com/pome-sh/digital-twins