Hi everyone!
Do you know if there is a way to provide feedback to an LLM after assertion?
I'm using promptfoo for checking my text -> graphql pipeline
I would like to validate generated query with python/js and provide an error back to LLM + regenerate the query
I read the docs, but couldn't find anything suitable for that.
Essentially, I'm trying to do smth similar to this:
https://www.promptfoo.dev/docs/guides/text-to-sql-evaluation/
but this example also doesn't have any sql validation + regeneration with feedback