What it does
The same design runs for all three primitives and records one row per item, framing and repeat, so a question that changes its mind under rewording is visible before it ships.
Benchmarks & research
@vcjdeboer/jev-reliability
Pre-flights a System One question before you put its number behind an if, reporting a decision flip rate rather than a confidence score.
Bookmark: @vcjdeboer/jev-reliability
Keep this in your collection.
Leave a noteWhat would you try with this?
: @vcjdeboer/jev-reliability
Only you can see your notes.
Screenshot unavailable. Open the experiment ↗
How you can use it
Start by writing out a single rule your app must decide, like checking if a message is urgent. Collect several sample messages and draft two or three ways to reword your question. This preparation lets you check whether small wording shifts cause the answers to change unexpectedly.
A developer can load your test messages into the tool using an access key generated for TypeSafe. The tool tests the reworded questions repeatedly to show if the decisions hold steady. Keep in mind that running these tests costs money on your TypeSafe account.