This topic was closed Feb 2024.
I read through all of the discussions and wanted to see if anyone had any updates that they could offer.
In that discussion I am leaning toward beninsteon's suggestion:

I would recommend against anything that is cloud dependent/connected. Additionally, given the importance of what this is doing, zwave or even wifi are not ideal.

I am using a hard wired motorized valve (eg. https://www.amazon.ca/Motorized-Stainless-Polarity-U-S-Solid/dp/B06XWG6ZLS) controlled via a Global Cache GC-100 (IP2CC would work also https://www.amazon.ca/Global-Caché-Ip2Cc-P-Contact-Closure/dp/B002ZV8FVI), connected to my network via Ethernet.

Using programs/network resources in ISY to trigger valve closure when water detected where it shouldn't be.

Also, I am interesting in sensors. I currently use INSTEON 2852-222 Water Leak Sensor.
I had one failure using the Insteon sensor. I have a pump room where I boost water pressure from 30 psi to 70 psi.
I had a one way valve rupture and it flooded the pit of the pump room where the sensor was located so fast that it drown the insteon sensor.
I never received the warning email. It seems like I need to have the water sensor in a housing that would restrict water exposure.
Any thoughts would be welcome.

I have a zwave watercop valve. Zwave is generally quite reliable. A hardwired valve would need to be controlled by a contact closure, which would necessarily involve some communications (wired or wireless) between the eisy and a contact closure device.