phone-harness
let your agent control your phone
The Lens
By Erik Loyd, SaaS CEO and former COO/CFO of an AWS Premier Partner.
Updated Oct 2026
phone-harness hands an AI agent your phone. Claude Code, Codex or any shell-capable agent gets eyes (on-screen text with tap coordinates) and hands (taps, swipes, typing) on a real iPhone or Android, with no jailbreak and nothing installed on the phone. MIT licensed and free.
iPhone runs through the Mac's iPhone Mirroring window with Apple's Vision OCR, so it needs macOS Sequoia, and the window stays in front unless you opt into an experimental background mode. Android goes over adb from macOS, Linux or Windows and reads the accessibility tree. iPhone Mirroring still isn't offered in the EU, so EU users get Android or cloud.
No spare phone? Phone Harness Cloud rents hosted iPhones and Androids at 5 cents a minute per phone, with $5 of promo credit to start; dedicated phones are custom-priced. Solo tinkerers: free on your own device. Teams doing real mobile testing: Appium, Maestro and BrowserStack remain the serious answers.
The catch: you're giving an LLM your logged-in phone, and telemetry is on by default. It sends your script, helper-call arguments (including text you type) and up to 20,000 characters of output to PostHog. Run phone-harness config set telemetry false first.
Free vs Self-Hosted vs Paid
free self hosted paid cloudFree: The harness itself, MIT licensed, on PyPI as phone-harness. Driving your own iPhone or Android costs nothing, and a cloud account needs no card.
Self-hosted: Runs locally as a uv or pip tool. iPhone: macOS Sequoia or later with iPhone Mirroring paired, plus Accessibility and Screen Recording permissions; iPhone Mirroring is not available in the EU. Android: adb on macOS, Linux or Windows, over USB or Wi-Fi (Android 11+). Anonymous-labelled telemetry is on by default and includes the script text, helper-call arguments and up to 20,000 characters of output; turn it off with phone-harness config set telemetry false.
Paid: Phone Harness Cloud (open beta): hosted iPhones and Androids at $0.05 per minute per phone, billed per minute, boot time free. New accounts get $5 of promotional credit (about 100 minutes, non-refundable). Top-ups are $5 to $200 by card and purchased credit doesn't expire. Dedicated phones (reserved around the clock, multiple profiles per phone, your choice of proxies, invoicing) are custom-priced via a demo call.
Free on a phone you own. Cloud phones are 5 cents a minute per phone (about $3 an hour), with $5 of starter credit.
What to do by team size
- Solo
- free on your own phone; an experiment, not a dependency
- Small team
- free locally; cloud phones at 5 cents a minute when an agent needs its own device
- Medium team
- cloud pay-as-you-go or a custom dedicated-phone deal; Appium or Maestro for real test suites
- Large team
- dedicated phones on a custom contract, or BrowserStack / Sauce Labs for device testing at scale
Get tools like this every Wednesday
One featured tool, three on the radar. No fluff.
A low score is not a verdict on quality. Young and niche tools start low by design. How we calculate scores
License: MIT License
Use freely, including commercial. Just keep the license.
Commercial use: ✓ Yes
About
- Owner
- shawn pana (User)
- Stars
- 4,906
- Forks
- 510
Explore Further
More tools in the directory
rea
Reverse engineer anything with agents, from app behavior down to native binaries.
453.7k ★openclaw
Your own personal AI assistant. Any OS. Any Platform. The lobster way. 🦞
394.4k ★everything-claude-code
The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
301.6k ★