Jev Sift
Classify first. Read selectively. A portable agent plugin and MCP tool for batch text classification.
Added to Jevfast
How it uses Jev
Jev Sift accepts a query or typed questions with a batch of file paths, public webpage URLs, or supplied text. It reads or fetches each item, sends its content to Jev, and returns compact per-item relevance probabilities or typed judgments for the agent to inspect.
What Jev decides
Illustrative Jev requestExample answers · not a recorded Jev response · Source ↗
Question 1 · sector
YOUR APP
INSTRUCTION
Choose its primary sector.
STATE
Two supplied text items: company-a says “We build payroll software for hospitals”; company-b sells outdoor camping equipment.
JEV · CHOICE
- healthcare
- other
App workflow
- Submit a search intent
Provide a query or one to eight typed questions together with uniquely identified items.
- Load candidate content
Read local paths, fetch supported public webpages, or use supplied text; each item uses exactly one source field.
- Judge each item
Send item content and the query or typed questions to Jev and return results in input order.
- Inspect promising context
Use relevance results to decide which files or pages the calling agent should open; the tool does not execute supplied tool descriptions.
Why it is interesting
It screens candidate context before the main agent opens or reads it, and can evaluate a tool description without executing the tool or predicting its result.