User repository on GitHub
Bridge that lets Claude Code verify an iOS app on a simulator, with TypeSafe's Jev choosing each step (planning stage)
- Forks
- 0
- Last pushed
- Updated Oct 2026
Repo stats from the GitHub API, cached Oct 2026.
Jev directory
CommunityJev Integrations & Adapters
Jev Ios Bridge
A bridge and Claude Code plugin that runs scripts of taps, typing and swipes on iOS simulators or Android devices and has Jev judge each checkpoint's screen text, returning passed, failed or inconclusive with screenshots and a log. Saved scripts rerun from the terminal with no Claude in the loop.
- Stars
- 3
- Language
- TypeScript
- License
- MIT
- Reading time
- 3 min
- Published by
- Community
- Author
- hugues-vnsgn
- Last verified
- Oct 2026
- Languages
- TypeScript
- Capabilities
- Open sourceRequires API key
Highlights
- Checkpoints like 'The order total reads $5' are judged by Jev reading the screen text; a check costs under $0.0002 in Jev tokens.
- iOS runs through MobileBuildMCP on a simulator; Android runs through adb and the bridge's own copy of mobilecli's device agent.
- Scripts work for SwiftUI, UIKit, Compose Multiplatform, Jetpack Compose and Android Views apps, and a saved script reruns from the terminal with no Claude.
- Version 1.4.0 adds experimental observed iOS actions, but the planned live simulator acceptance was not completed.
- Claude writes and runs checks through the plugin's /jev-ios-bridge:test-ios and /jev-ios-bridge:test-android skills.
Quickstart
bash
claude plugin marketplace add hugues-vnsgn/jev-ios-bridge
claude plugin install jev-ios-bridge@jev-ios-bridgeWatch out
README reviewed 2026-10-17; v1.4.0 shipped with code/package checks passed and the live simulator acceptance not completed, in driven mode Jev made 10 of 13 actions in one live trial, and the README says a run can be slower than letting Claude drive the device directly.
Questions about this entry
Short answers drawn from the linked project or page.
A bridge and Claude Code plugin that checks an iOS or Android app the way a person reads its screen. You write a script of taps, typing, swipes and waits with checkpoints, and Jev reads the screen text and judges each claim, giving one verdict: passed, failed or inconclusive.
At each checkpoint the bridge sends the screen's text and your claims to Jev, TypeSafe's decision model, which returns probabilities. A whole driven run of 13 decisions and 2 checkpoints used 54,000 to 58,000 input tokens and cost about $0.002.
For v1.4.0: a Mac, Node 24 or later, a TypeSafe API key and Claude Code. iOS needs Xcode and a simulator runtime; Android needs Android Studio, adb and a device on Android 12 (API 31) or later. Screen text goes to TypeSafe at each Jev call.
TypeSafe charges $0.042 per million input tokens for jev-1.13.0 and output is free. One checkpoint costs under $0.0002, one driven decision up to $0.00025, and the README's whole driven run about $0.002. Claude's share of a run is usually larger.
Yes. A version 1 script reruns from the terminal with no Claude at all. Claude writes the script through the plugin, and the evidence goes to .jev-runs/<run-id>/ with screenshots, a report and a log.
On iOS it drives simulators, not real iPhones; on Android real phones are supported but untested. Jev's judgment covers English screens, a run can be slower than direct driving, and screen text is sent to TypeSafe at each checkpoint and driven decision.
A do step names what should be achieved and the text that shows afterwards, and Jev picks each action. It is off by default and enabled in two places: drivenMode in .jev/config.json and the plugin's Experimental driven mode setting or JEV_EXPERIMENTAL_DRIVEN=1.
Reactions & coverage
Posts, threads, and videos about this entry from around the web.
YouTube: Ray Amjad puts Jev inside an agentic coding loop and looks at what it costs to run.
YouTube: RoboNuggets puts Jev to work inside Claude Code.
Reddit: A Claude Code mod replaces the auto-mode classifier with Jev, 2x faster.
More Jev resources like this
A Claude Code mod that sends each prompt to Jev through OpenRouter and turns the returned 0-3 score into a reasoning effort level, from low to xhigh. Shadow mode is on by default, so it only logs what it would set until you turn it off.
Jev Integrations & Adapters#community#github#claude-code
Community
gebeer/jev-effortTypeScript
gebeer/jev-effortTypeScriptA local stdio MCP server from bgivenb that runs whole browser and macOS goals for your agent: your agent supplies the goal, exact values, and completion conditions, and Flick loops observe, decide, and act with TypeSafe Jev. TypeScript core with Playwright and Swift drivers.
Jev Agents & MCP#community#github#mcp
Community
bgivenb/flick-computer-useTypeScript
bgivenb/flick-computer-useTypeScriptAn unofficial Claude Code plugin that checks 'done' claims and small judgement calls with TypeSafe Jev and keeps receipts: its done-gate can reject 'all tests pass' and ask for a real check, so evidence beats eloquence.
Jev Integrations & Adapters#community#github#claude-code
Community
ismaildasci/claude-refereeTypeScript
ismaildasci/claude-refereeTypeScriptA Claude Code mod that asks Jev one Noul question per skill for every prompt, ranks the skills above the threshold and passes them to the model as a suggestion. It never loads a skill itself, and a timeout, outage or missing key only skips the routing.
Jev Integrations & Adapters#community#github#claude-code
Community
gecm0/skill-router-modTypeScript
gecm0/skill-router-modTypeScriptA Playwright-based end-to-end test runner for web apps where Jev powers plain-English assertions, actions by description, goal-driven runs and failure triage. Its API mirrors Playwright Test, with test, expect, fixtures and page objects, and every Jev feature is optional.
Jev Tools & CLIs#community#github#testing
Community
neeraj719714/intelliwrightTypeScript
neeraj719714/intelliwrightTypeScriptA fork of codex-with-chatgpt that adds TypeSafe Jev plus deterministic rules to decide when Codex hands planning, debugging or review to web ChatGPT and when it switches back. Smart switching is off by default, and with it off the tool behaves exactly like the original.
Jev Integrations & Adapters#community#catalog-import#logicrw
From 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.