Skip to content
JevHub

[ Built with Jev · Coding agents ]

jev-harness: propose, ask, decide

A research-stage coding harness where an LLM proposes and Jev reviews.

By TypeSafe Community (@TypeSafeAI on GitHub) ·

jev-harness

TypeSafeAI

GitHub preview of TypeSafeAI/jev-harness

How it uses Jev

An LLM proposes one action, Jev answers four narrow questions about it, and code produces evidence for the host to act on. Maintained by a community organisation independent of TypeSafe.

Summary written by JevHub. Figures are the author's own; we haven't reproduced them. The project belongs to its author.

More builds

Roshan Chandna

@roshanchandna

Your coding agent is one bad instruction away from running rm -rf ~/.aws. I built Auto-Guard with Jev (@typesafeai) to check every tool call before it runs. Here it stops a credentials delete but lets the build cleanup through. It's open source! Install in the replies 👇

24likes on X

jev-review

devagrawal09

GitHub preview of devagrawal09/jev-review

A code-review workflow for diffs or whole codebases, with a quiet local dashboard.

stanley-code

devagrawal09

GitHub preview of devagrawal09/stanley-code

A command-line tool for checking and understanding code changes, described in plain words.