GitHub · Coding and context · Tools and apps
dzhng/jevgrep
Seeing what a code-search step looks like when the decision model does the picking.
What it does
A CLI that finds code by asking what it does, built for coding agents: Jev picks the files and the source context worth reading instead of the agent grepping its way in. The README's headline is a cost claim — "Same intelligence. About 30% lower cost." — measured on their own ten-task SWE benchmark.
The repository's own description: “Find code by asking what it does. A CLI for coding agents that uses Jev to discover relevant files and source context.”
How it works
A TypeScript monorepo (apps/, packages/, evals/, specs/, skills/, docs/) managed with bun, and published to npm as @dzhng/jevgrep. The repo carries its own eval suite, which is where the ten-task benchmark lives — the number is in the repository, not only in the README.
The repository, by the numbers
Read from the GitHub API on 2026-10-09. Stars and the last push move daily — quote them with the date, the way we do.
What we checked
- the README's headline, its npm badge and the Node 22+ requirement
- the top-level tree: apps, packages, evals, specs, skills, docs, plus bun.lock
- GitHub metadata on 2026-10-09: MIT, TypeScript, 2,482 stars, last push 2026-10-02
What we did not check. We have not run the CLI or its evals, so the ~30% figure is the author's, on a ten-task benchmark we did not reproduce.