What it does
Reviews one proposed coding action with narrow Jev questions before deterministic code prepares evidence for host authorization.
Agent tooling
jev-harness
Bookmark: jev-harness
Keep this in your collection.
Leave a noteWhat would you try with this?
: jev-harness
Only you can see your notes.
Screenshot unavailable. Open the experiment ↗
How you can use it
A developer can test how an outside AI service called Jev reviews code changes. The software asks the AI four specific questions about a proposed action. It prepares these answers as evidence for a person to review. The system only records the proposed code changes and never actually applies them.
You can try the included web test to compare how different tools handle example tasks. This step requires an access key that connects the application to TypeSafe. Your web browser saves your recent test results. It can hold up to thirty runs before running out of space.