Does my code get sent to a model, or does this stay local?

It leaves your machine. analyze sends changed code and nearby context to TypeSafe and OpenAI. What stays local: the extension talks only to the local server on 127.0.0.1:4731, and nothing is posted to GitHub, since the tool is meant for reviewing your own agent's PRs locally.

Does it review the whole diff, or just the first part?

Only the first 12 change units, meaning diff hunks, in path order; --max-units raises the cap to 100. Files whose change went unanalyzed keep GitHub's original code, and both the CLI and the popup report how many were left out.

Will it work on GitHub's new pull request UI?

Not yet: the classic Files changed page is supported and the new /changes page is not, so you have to switch back under your profile picture and Feature preview. Separately, if the PR gets newer commits the extension keeps GitHub's code until you rerun analyze and click Refresh report, so a stale report is never shown.

Do P0, P1 and P2 mean a priority is a correctness verdict?

No. The README says priorities mean attention, not correctness: P0 files open by default and P1 and P2 start collapsed, so a huge agent diff stays readable. The bands are configurable by copying config/policy.json to .jev-reviewer.json in the reviewed repo, or by passing --policy.

Where do my API keys live, and how safe is that?

jev-reviewer setup writes them to ~/.config/jev-reviewer/credentials.json, owner-only but not encrypted, and the README prefers that over exported environment variables, which a coding agent's shell often cannot see. Replace an OpenAI key later with node scripts/setup-keys.mjs --replace-openai rather than editing the file by hand.

What has to be true before it can review a pull request?

The GitHub CLI signed in with gh auth login, a local clone whose origin is the PR's repository, and the local server left running. The sequence is npm link once, then jev-reviewer setup, jev-reviewer doctor to check tools, key sources, the server and provider access, jev-reviewer serve, and finally jev-reviewer analyze --pr URL --repo /path/to/clone.

Can I see it work before spending anything on provider calls?

Yes. With Node.js 22 or newer, npm install and npm run demo serve a replay at http://127.0.0.1:4731/demo that makes no provider calls at all. The bundled recording is honest about that: it uses real Jev classifications with labelled, prepared explanation copy. Optionally, uv tool install graphifyy adds a local code graph for better context.

CommunityJev Tools & CLIs

Jev-Code-Reviewer

A local Chrome extension plus CLI that classifies each agent PR hunk as P0, P1, or P2 with Jev, showing P0 by default and human-readable diff summaries.

See highlights

Open github.com

Stars: 140
Language: JavaScript
License: MIT
Published by: Community
Author: egma-ai
Last verified: Sep 2026

egma-ai/jev-code-reviewer

Organization repository on GitHub

View on GitHub

Review behavior, not just diffs. Jev prioritizes human attention; OpenAI explains the changes. Local CLI + agent skill + GitHub extension.

Forks: 2
Last pushed: Updated Oct 2026

Repo stats from the GitHub API, cached Oct 2026.

About this project

Most PRs that are getting generated by agent hard for human mind to comprehend as it shows up with lot of file changes at once. This is an attempt to reduce the mental burden by classifying each change in a review to P0, P1, P2. Only P0 are shown by default. The priorities are configurable. The diffs are also show using a natural language. The original code is one toggle away. It runs on your computer, for reviewing your own coding agent's PRs, and posts nothing to GitHub.

Quoted from the project's README. Read the full README.

Tagscommunity code-review typescript

Works forJev for Classification Jev for Code

Used inJev for Software & DevTools

Languages: TypeScript
Platforms: Chrome extension CLI
Capabilities: Open sourceRequires API key

Highlights

Node.js 22+. npm run demo serves a replay at http://127.0.0.1:4731/demo without provider calls.
jev-reviewer analyze --pr needs gh auth and a local clone; results stay off GitHub.
Chrome extension loads unpacked from extension/; jev-reviewer serve must stay running.
Priorities are configurable; default view hides P1 and P2 to shrink huge agent diffs.

Watch out

MIT. Needs TypeSafe and OpenAI keys in setup for live review; demo recording uses prepared copy.

Questions about this entry

Short answers drawn from the linked project or page.

Reactions & coverage

Posts, threads, and videos about this entry from around the web.

Hacker News: Jev classifies agent PR hunks as P0, P1, or P2 so reviewers see P0 first.

namanbhulawat47 pointsJev Based Code ReviewRead the thread on Hacker News

Hacker News: Thread on Selmar's Jev code-review calibration notes and ordering.py comparisons.

Lectem2 pointsCalibrating Jev as a Code ReviewerRead the thread on Hacker News

X: Magic Jev Ball for code review

X: jev-review: a local-first score loop for coding agents

More Jev resources like this

GenOffice

9.2k

An open-source AI office suite whose local file search reranks the top 20 FTS hits with TypeSafe Jev relevance scores before showing results.

Jev Tools & CLIs#community#typescript#desktop

Community

genspark-ai/genofficeTypeScript

jevgrep (code search)

2.6k

A CLI that asks Jev which files and declarations answer a repository question, then prints verbatim excerpts for a coding agent to use.

Jev Tools & CLIs#community#typescript#cli

Community

dzhng/jevgrepTypeScript

Pi TypeSafe Router

A Pi coding-agent extension that uses Jev to classify each request and route it to one of three configured models. Supports TypeSafe, Cloudflare AI Gateway, Vercel AI Gateway and OpenRouter backends, with doctor checks, shadow mode and an off-by-default setup.

Jev Tools & CLIs#community#coding-agents#routing

Community

jekozyra/pi-typesafe-routerTypeScript

Sedum

Open-source, pre-alpha CLI that runs plain-English browser tests on Playwright, using Jev for bounded decisions: resolving each step, picking elements among offered candidates and judging verify claims with probabilities. Bring your own TypeSafe key; MIT licensed.

Jev Tools & CLIs#community#e2e-testing#playwright

Community

sedum-dev/sedumTypeScript

jev-lint

Fuzzy linter hook for Claude Code and Codex: after the agent edits a file, it sends the added code to Jev, which answers Noul checks against team rule packs in about 0.3 seconds, letting the agent fix violations before code review.

Jev Tools & CLIs#community#linter#claude-code

Community

ckorhonen/jev-lintHTML

semcheck

Go linter whose rules are plain-English questions: AST matchers select nodes and Jev answers one Noul per fragment above the rule's threshold. It runs standalone, under go vet, and as a golangci-lint plugin, with cached answers so runs only ask about changed code.

Jev Tools & CLIs#community#linter#go

Community

arturobermejo/semcheckGo

From the Jev guides

Original write-ups on what this entry is for.

Jev on the Gateways: Vercel, Workers AI, and OpenRouter

Oct 12, 2026

The request shape barely changes across surfaces; the surrounding contract does. When to call the API directly and when a gateway, a Worker binding, or a router entry earns its extra hop.

Jev Tools & CLIs#gateway#vercel#cloudflare

Read article

Exporting Decision Models: GGUF, ONNX, and Device Runtimes

Oct 10, 2026

Taking a decision checkpoint out of Python: what the GGUF and browser WebGPU exports preserve, what they change, and the parity evidence the model cards publish.

Jev Tools & CLIs#open-models#local#onnx

Read article

Records at Rest: Documents, Inboxes, and Risk Queues

Sep 28, 2026

Jev labels a record that already exists. Code moves, flags, or holds it. How the document, mail, and transfer tools keep the action out of the model.

Jev Tools & CLIs#classification#email#security

Read article

Back to all resources

From the Jev community

Posts from builders shipping with Jev right now.

Follow @typesafeai