Native phone app
boboabw
Phone Agent
An iOS app that combines shared-screen observation with an ESP32-S3 Bluetooth accessory.
- Hardware
- ESP32-S3; default firmware uses GPIO48 RGB
- Runs on
- Native iPhone app
A FIELD GUIDE TO PHONE AGENTS
One phone.
More ways to build.
Explore the projects connecting ESP32 hardware and iPhone agents. See where the intelligence runs, how the screen is captured, and what actually sends the input.
11source-linked projects
05architecture routes
03languages
01 / INDEX
Different projects. Different approaches.
A starting selection across native apps, browser agents and computer-hosted systems. Reading order is not a performance ranking.
Browser agent
cccat6
iPhone Agent Harness
Safari hosts the agent; a local WebRTC link sends actions to a Bluetooth ESP32 bridge.
- Hardware
- ESP32 C2, C3, C5, C6, S3 or classic; board requirements vary
- Runs on
- Safari on iPhone / iPad
Computer-hosted
getGate
Oh My Phone Use
A Mac MCP setup for iPhone screen control and a separate Bluetooth hands-free calling mode.
- Hardware
- ESP32-PICO-KIT-1 with Classic Bluetooth
- Runs on
- Mac MCP server plus iOS app
Native phone app
ja1chen
Actell / 执言
An experimental CoreDevice agent with optional ESP32-S3 USB networking.
- Hardware
- Optional ESP32-S3 with native USB for NCM
- Runs on
- Native iPhone app with Swift / Rust backend
Native phone app
onehumanbeing
ESP32 iPhone Agent / Handsy
A native iPhone agent with board-side AirPlay reception and USB mouse input.
- Hardware
- ESP32-S3 with PSRAM and native USB
- Runs on
- iPhone app; computer for building and signing
Native phone app
ZhaoYangGG
C3 Agent
An on-phone AutoGLM-inspired task loop with ESP32-C3 Bluetooth input.
- Hardware
- ESP32-C3; USB is for power and flashing
- Runs on
- Native iPhone app
02 / ROUTES
Start with the architecture.
The board is only one part. Screen capture, agent execution and input can live in different places.
01
Native phone apps
The agent loop lives in an iPhone app; board roles range from input to video relay.
02Browser-based agents
Safari handles the conversation and screen sharing; a local bridge delivers input.
03Computer-hosted setups
A Mac, Linux or Windows host runs the agent, receiver or external MCP connection.
04Input building blocks
Firmware supplies actions, while another client must observe the screen and decide.
05Voice and call bridges
Bluetooth audio links a phone to a voice agent. This does not imply screen control.
03 / SOURCES
A directory with receipts.
Every entry names its maintainer, links to the reviewed source revision and separates author claims from our own verification. No hardware benchmarks are claimed.
How we select and review04 / NOTES
Looking for the existing tutorials?
The original setup guides are still here, clearly scoped to ESP32 iPhone Agent by Henry Yuan. Other projects have their own requirements.
Open the project build notes