Sometimes the Claygent provides bad leads when asking it do research, but the fix to this is...
- 1.
- 2.
Zenrows scrape of source (or ideally the original input so it doesn't cost as much)
- 3.
I would love a baked in extra validation model in the Claygent (like two in one) that is like, "your only job is to verify what the first Claygent found and here is all the context and remove the bad stuff."