User repository on GitHub
Browser automation CLI for AI agents, powered by the Jev model's millisecond decisions and near-zero inference costs
- Forks
- 1
- Last pushed
- Updated Sep 2026
Repo stats from the GitHub API, cached Oct 2026.
Jev directory
CommunityJev Tools & CLIs
Jev Browser — mrlyk
A browser automation CLI for AI agents: drive Chrome with natural language in an interactive terminal or structured commands like jevb page act, signed in with a TypeSafe or OpenRouter key. Prebuilt packages target macOS Apple Silicon; Node.js 22+.
- Stars
- 5
- Language
- Rust
- License
- APACHE-2.0
- Reading time
- 3 min
- Published by
- Community
- Author
- mrlyk
- Last verified
- Oct 2026
This is one of several Jev listings for the same project. The canonical entry is Jev Browser — jkudish.
- Languages
- Rust
- Platforms
- CLIBrowsermacOSOpenRouter
- Capabilities
- Open sourceRequires API key
Highlights
- Install with npm install -g jev-browser-cli; the CLI finds or downloads Chrome on first page open, and prebuilt packages support macOS Apple Silicon only.
- Commands follow jevb resource action session; the README chains page open, page act and page snapshot on one named demo session for agent and script use.
- The TUI offers /connect (existing Chrome, CDP, headed, headless or session), /tab, /scroll and /status; /exit keeps the browser running for reconnect.
- Sign in with jevb auth login using a TypeSafe or OpenRouter key; environment keys override saved ones and TypeSafe wins when both are present.
- A companion agent skill installs with npx skills add Mrlyk/jev-browser --skill jev-browser, and each page act performs a single operation.
Quickstart
bash
npm install -g jev-browser-cli
jevb page open demo https://www.baidu.com --headed
jevb page act demo "Search for jev"Watch out
Copy reviewed against the repo README on 2026-10-15; the captured text is truncated from the agent-commands section. The prebuilt package supports only macOS Apple Silicon and other platforms need a source build, and no latency or success-rate benchmarks are published.
Questions about this entry
Short answers drawn from the linked project or page.
A browser automation CLI for AI agents powered by Jev: control Chrome with natural language in an interactive terminal, or call structured commands like jevb page open, jevb page act and jevb page snapshot from an agent or script. The npm package is jev-browser-cli.
npm install -g jev-browser-cli, then use jevb or jev-browser. It requires Node.js 22+, and the prebuilt package supports macOS Apple Silicon only; other platforms need a source build. When you first open a page, the CLI finds Chrome or downloads it if needed.
Running jevb opens a TUI bound to one tab, with commands like /connect, /tab, /up, /down, /open, /status and /help. /connect auto attaches to Chrome, headed or headless launches a new browser, and /connect cdp 9222 targets a specific address.
TypeSafe or OpenRouter. If no key is configured, jevb prompts you to sign in and input is hidden; you can also run jevb auth login or jevb auth status. Environment variables TYPESAFE_API_KEY or OPENROUTER_API_KEY override saved keys, and TypeSafe wins when both exist.
The README provides a copy-paste prompt: install the CLI with npm install -g jev-browser-cli, run npx skills add Mrlyk/jev-browser --skill jev-browser, select your agent, then point it at the skill file for browser tasks in the project.
Each page act performs one operation; a search may include typing and pressing Enter. When you already know the URL, selector or key, the README suggests deterministic commands such as page open, element click or keyboard press that skip the model decision entirely.
The selected tab stays bound to the session, and /exit keeps the browser running so you can reconnect with the printed session command. /exit --close also closes a browser created by that interaction. The previous browser stays open when /connect starts a new one.
Reactions & coverage
Posts, threads, and videos about this entry from around the web.
X: Text-to-screenshot (Malay Vasa)
X: Google's MediaPipe Decision Maker, tested in the browser
X: Playwright CLI + Jev vs Playwright MCP
X: jev-ultrafast splits every browser decision into operation, target and value. JEV passed 18,000 stars in its…
More Jev resources like this
Remote control for coding agents: give Claude Code, Codex, Cursor, OpenCode, Pi and others a task from your phone, and get a notification when the result is ready while your laptop stays closed. Self-hosting is a free MIT install; the hosted version has a free plan and paid plans from $19/mo.
Jev Tools & CLIs#community#catalog-import#logicrw
Community
bennykok/omg.devTypeScript
bennykok/omg.devTypeScriptRust coding-agent harness (MIT, ~20k stars) whose crates/jcode-base/src/jev.rs is a typed TypeSafe Jev decision transport, used for memory pruning (memory_jev.rs), browser navigation (browser_jev.rs), and voice routing (voice_intent.rs) with separate provider selectors per purpose.
Jev Tools & CLIs#community#rust#coding-agent
Community
1jehuang/jcodeRust
1jehuang/jcodeRustA Mac-only voice assistant: say Hey Jev or hold right Option and Jev decides every command in one call for $0.00004 per request, while Whisper transcribes locally, Fish Audio S2.1 Pro speaks replies, and an LLM wakes only for questions.
Jev Tools & CLIs#community#catalog-import#logicrw
Community
henryklunaris/hey-jevPython
henryklunaris/hey-jevPythonOpenChamber is an MIT-licensed workspace built on the OpenCode agent for running and reviewing AI coding work on desktop, web, VS Code, and mobile, with session goals, multi-run comparisons, and private remote access.
Jev Tools & CLIs#community#catalog-import#logicrw
Community
openchamber/openchamberTypeScript
openchamber/openchamberTypeScriptVercel Labs' Unix-like coding agent CLI written in Zig, shipping a typesafe_permission_reviewer builtin that runs the agent's permission decisions through Jev rather than an LLM call. Model-agnostic and embeddable via libfx and ACP.
Jev Tools & CLIs#community#coding-agents#cli
Community
vercel-labs/fxZig
vercel-labs/fxZigVellum Assistant is an MIT-licensed personal AI assistant with eight memory types, hourly proactivity, and permission-gated computer use, reachable from macOS, Windows, iOS, web, voice, email, Telegram, Slack, and Twilio.
Jev Tools & CLIs#community#catalog-import#logicrw
Community
vellum-ai/vellum-assistantTypeScript
vellum-ai/vellum-assistantTypeScriptFrom the Jev guides
Original write-ups on what this entry is for.
From the Jev AI model community
Posts from builders shipping with Jev right now.