A taxonomy of nine jobs found across the catalogue, from corpus classification and queue triage to agent routing, guardrails, context pruning, database judgments, and UI ranking.
Original by Made with JevClassificationMade with Jev guideSource reviewed
Before you dive in
What you’ll find in the original
Classifying a large corpus is compelling when the task is a repeatable closed question rather than open-ended synthesis.
Agent loops benefit when Jev selects among legal next actions and the host executes the chosen branch.
Guardrails, pruning, and ranking still need explicit thresholds, source data, and deterministic handling around the model.
Worth knowing
Counts and medians describe the catalogue snapshot; individual performance numbers are claims from each linked builder.