Google web risk API caching

See more:

Google

I have written a small console app to integrate with Google Web Risk API which allows me test if a url is malicious or not.

I have a question about caching. If I want to build my own result caching solution, does anyone know if I maintain cache entry for the whole url or the just cache by domain name is suffice?

For example

mysite.com/badpage.html

Will Web Risk API mark the whole mysite.com as malicious? In which case I just have to store the host mysite.com has malicious?

What I have tried:

Write a small console app integrating with the google web risk api.