Jev-Code-Reviewer
Local PR reviewer that has Jev prioritise each change P0-P2 so humans read the important hunks first, with plain-language explanations.
Overview
Jev-Code-Reviewer addresses the burden of reviewing agent-generated pull requests with many changed files at once. It classifies each change into P0, P1 or P2 with Jev, showing only P0 by default (priorities are configurable), and uses OpenAI to explain the diffs in natural language, with the original code one toggle away. Per the yibie list, it asks Jev for a priority score per changed unit and computes a priority gap that a local uncertainty policy turns into the order a human should read hunks.
It runs on your own machine for reviewing your own coding agent's PRs and posts nothing to GitHub. It combines a local CLI and server, an agent skill and a Chrome extension that overlays results on a PR's Files changed page after pairing with a token. Setup needs Node 22+, the GitHub CLI signed in and a local clone, and stores TypeSafe and OpenAI keys; a doctor command checks tools and access. A demo replays real Jev classifications without provider calls.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Choice | What review priority does this change deserve? | P0 · P1 · P2 |
| Score | How important is this changed unit for human review? | — |
How it works
- Analyze a PR from a local clone
- Jev scores each changed unit's priority
- Local policy orders hunks by priority and uncertainty
- OpenAI explains surfaced changes
- Chrome extension shows prioritised view on GitHub
Strengths
- Focuses human attention on high-priority changes
- Posts nothing to GitHub
- Demo works offline
Caveats
- Needs both TypeSafe and OpenAI keys
- Unpacked Chrome extension and local server setup
Getting started
`npm install && npm run demo`, then open http://127.0.0.1:4731/demo; for real PRs use `jev-reviewer setup` and `jev-reviewer analyze --pr <url> --repo <path>`.At a glance
- Type
- CLI
- Status
- prototype
- Language
- TypeScript
- Runtime
- Node CLI and server, Chrome extension
- Has code
- yes
Use cases and integrations
How curated lists describe it
Code review: asks Jev for a priority score per changed unit and returns a `priorityGap` that a local uncertainty policy turns into the order a human should read the hunks in, while OpenAI explains the ones that surface.
yibie/awesome-jevReview behavior, not just diffs. Jev prioritizes human attention; OpenAI explains the changes. Local CLI + agent skill + GitHub extension.
hellogumbo/awesome-jevReviews your own agent's pull requests on your machine and posts nothing to GitHub, prioritising where a human should look rather than restating the diff.
MrJev/awesome-jev
Frequently asked questions
What does Jev decide?
The review priority (P0-P2) of each change.
Does it comment on GitHub?
No, it runs locally and posts nothing.
Who explains the changes?
OpenAI writes the plain-language explanations; Jev only prioritises.
Related projects
CLI that scores every test against a git diff with Jev and emits filter arguments for vitest, jest, node, Playwright, cargo, go and rspec.
ProjectbetaTypeScriptOh My Pi skill for method-first bulk classification: decide when to use Jev, then pilot and freeze a rubric for judge_batch.
Agent skillbetaPythonMCP server that finds relevant code with Jev Choice and Noul checks instead of grep sweeps, and audits diffs against task intent.
MCP serverprototypePythonCLI, MCP server and Claude Code plugin for closed-form yes/no, choice, score and rank decisions via Jev, OpenAI Decisions or Claude.
MCP serverbetaTypeScriptHermes Agent plugin exposing Jev as four coding-workflow tools: jev_check, jev_route, jev_score and jev_evaluate.
PluginbetaPythonSemantic code search and plain-English review rules that send repository chunks to Jev and return scored source locations.
ProjectbetaTypeScript