What it does
The default target score is based on a simple three-level scale, and there is no proof that reaching this score guarantees better final results.
Agent tooling
Jev Prompt Enhancer
An agent skill scores a task prompt, asks the host coding agent to revise its weakest part, and rechecks it up to three times.
Bookmark: Jev Prompt Enhancer
Keep this in your collection.
Leave a noteWhat would you try with this?
: Jev Prompt Enhancer
Only you can see your notes.
Screenshot unavailable. Open the experiment ↗
How you can use it
Ask your coding assistant to use the prompt-enhancer skill with a task, such as fixing a bug or reviewing a change. It sends your instructions to Jev to identify unclear context, constraints or expected output. The assistant needs a Jev service access key before it can perform that check.
The assistant revises the weakest part, rechecks it and shows the resulting prompt before continuing. It stops at the target score or after three attempts. The score comes from a three-level rubric; meeting the target is not evidence that the final coding result will be better.