TypeSafe Jev Tweet Classifier
TypeSafe Jev acts as a pre-submission classifier for tweets, returning probabilities on posting worthiness, news/personal categorization, and twin hook risk, allowing the user to decide whether to proceed with posting.
Most people ask a chat model if a tweet is good. I stopped.
TypeSafe Jev sits in front of my Buffer queue now. One candidate goes in. Three typed questions come back in the same call: worth posting, news or personal, twin hook risk. My script reads the probabilities and either skips or lets me write.
Useful part: Jev cannot draft the tweet. That is the feature. Decisions stay in code. The voice stays mine.