physiclaw 0.0.10

Give AI agents a physical finger to operate any phone — robotic stylus + camera, no APIs needed.

PhysiClaw

What if an AI agent could use your phone — just like you do?

PhysiClaw gives AI agents an eye (camera) and a finger (robotic arm) to physically operate any phone. It looks at the screen, decides what to do, and taps.

Order food delivery. Check your email. Shop for groceries. Book a hotel. Any app, any phone, iOS or Android.

No OAuth tokens. No ADB cables. No APIs. No app to install. No developer setup. Just unlock your phone, put it on the desk, and let the agent work.

The tradeoff? PhysiClaw needs hardware: an embedded system running GRBL/grblHAL firmware to control a gantry (X/Y) and stylus (Z), plus a USB camera. A compact desktop rig that gives your AI agent physical presence.

Quickstart

macOS only for now. Hardware bill-of-materials below.

# 1. Install the CLI (uv + Python 3.12 + physiclaw, all isolated under ~/.local/bin)
curl -fsSL https://raw.githubusercontent.com/echosprint/PhysiClaw/main/install.sh | bash

# 2. Check your environment
physiclaw doctor

# 3. Download the local vision model (~100 MB, one-time)
physiclaw setup local-vision-model

# 4. Plug in the GRBL arm + USB camera, then start the server
physiclaw server                 # leave running in one shell

# 5. In another shell — interactive arm/camera calibration
physiclaw setup hardware

Then point your MCP client (Claude Desktop, etc.) at http://localhost:8048/mcp.

How It Works

 Camera ──→ AI Agent ──→ 3-Axis Arm ──→ Camera ──→ Aligned?
 (read screen) (decide)   (move stylus)  (check pos)   │
      ▲                                            Yes │ No
      │                                             │  │
      │       Touch Phone ◄─────────────────────────┘  │
      │            │                                    │
      └────────────┘ (next action)       adjust & retry ◄┘

One camera, two modes:

• Park + screenshot — stylus parked away, clear view of the full screen to read content
• Screenshot — stylus visible in frame, check its position relative to the target
• Stylus moves on X/Y to reach any point, up/down (Z) to touch or release

The loop is simple: look → think → move → confirm → touch → repeat.

Why PhysiClaw

Today's AI agents can control your computer — but they hit walls everywhere:

• Want to order food? Need a delivery API + OAuth.
• Want to check your bank? Blocked by data walls.
• Want to book a ride? Another service integration.
• Every new skill/service = new OAuth, new API, new setup. Tedious, fragile, limited.

PhysiClaw takes a different approach: let the AI agent physically use your phone. A camera sees the screen. A robotic finger taps it. No OAuth to apply for. No API to integrate. No app can detect or block it — because to the phone, it's just a finger.

One setup. Every app. Just put an unlocked phone on the desk.

System Architecture

┌───────────────────────────────────────┐
│           AI Agent (Brain)            │
│  Claude Desktop / OpenClaw / etc.     │
│  Sees screen → decides → calls tools  │
└──────────────────┬────────────────────┘
                   │ MCP Protocol
                   ▼
┌───────────────────────────────────────┐
│     PhysiClaw MCP Server (Python)     │
│                                       │
│  Tools:                               │
│   · screenshot       (camera)         │
│   · park             (retract)        │
│   · move             (X/Y plane)      │
│   · tap / swipe      (Z down + move)  │
└──────────┬────────────────┬───────────┘
           │                │
       USB Camera     USB Serial (GRBL)
           │                │
           ▼                ▼
    ┌────────────┐   ┌───────────────┐
    │   Camera   │   │ GRBL Board    │
    │  (above)   │   │ (embedded)    │
    │            │   │ X/Y gantry    │
    │            │   │ Z stylus      │
    └────────────┘   └──────┬────────┘
                            │ touch
                            ▼
                   ┌─────────────────┐
                   │  Phone          │
                   │  (unlocked)     │
                   └─────────────────┘

Hardware

Bill of Materials

Component	Item	Qty	Est. Price
GRBL Arm	Paixi Kuaichaobao pen plotter P25 (X/Y gantry + Z servo)	1	~$80
Camera	UGREEN 1080P USB camera, fixed focus	1	~$14
Stylus	Capacitive stylus, conductive fiber tip 8-10mm	1	~$1.5
Camera mount	Gooseneck desk clamp, metal, 50cm	1	~$2
Phone mount	Anti-slip pad + L-shaped blocks	1 set	~$1.2
USB Hub	USB 3.0 Hub (extend Mac USB ports)	1	~$13
Total (excluding computer)			~$112

Camera Setup

• Camera: straight above the screen center, ~25cm distance, reads screen content and checks stylus position

Phone Mounting

• Phone placed face-up flat on the arm platform
• Anti-slip pad + L-shaped blocks for positioning, ensuring consistent placement

Communication Protocol (PhysiClaw ↔ GRBL Arm)

GRBL G-code (USB → GRBL Arm)

All commands used in this project:

G91                    # Relative coordinate mode (default)
G0 Xxx Yyy Fxxx        # Rapid move on X/Y plane (position stylus)
G1 Xxx Yyy Fxxx        # Linear move at constant speed (swipe gesture)
M3 S12                 # Stylus down (touch screen)
M3 S0                  # Stylus up (release screen)
M5                     # Servo off
G90                    # Absolute coordinate mode (for park)
G0 X0 Y0 F5000         # Return to home position
$$                     # Query all GRBL parameters
?                      # Query real-time position

Protocol: USB serial (CH340, 115200 baud). Send one line at a time, wait for ok before next.

Key GRBL Parameters

Parameter	Meaning	Typical Value
$100 / $101	Steps per mm (X/Y)	80
$110 / $111	Max speed mm/min (X/Y)	5000
$120 / $121	Acceleration mm/sec² (X/Y)	200
$22	Enable Homing	1

MCP Protocol (MCP Client → PhysiClaw)

Tools communicate via stdio or SSE with JSON messages. MCP is a standard, language-agnostic protocol.

Tech Stack

Language

Python 3.12+

Key dependencies

(All installed automatically by install.sh — listed here for reference.)

• pyserial — send G-code to the GRBL board over USB serial
• opencv-python — USB camera capture
• mcp — MCP server framework
• rapidocr + onnxruntime — on-device OCR + icon detection
• httpx, typer, croniter — runtime, CLI, scheduling

No Anthropic SDK needed — Claude (or any other LLM) runs on the MCP client side.

Platform Compatibility

Mac / Windows / Linux (Raspberry Pi) all supported. The only platform difference is serial device names (Mac: /dev/tty.usbserial-xxx, Windows: COM3, Linux: /dev/ttyUSB0).

Code Structure

physiclaw/
├── server.py         # MCP Server entry point, exposes tools
├── core.py           # Central orchestrator (arm + camera + calibration)
├── arm.py            # GRBL G-code controller (tap, swipe, move)
├── camera.py         # USB camera capture and green flash detection
├── vision.py         # YOLOX phone detection and camera discovery
├── calibrate.py      # 5-phase calibration workflow
└── grbl.py           # Auto-detect GRBL serial port

Operation

The AI agent does not output coordinates — only direction and distance level. Each step is verified by photo.

Directions: up / down / left / right / up-left / up-right / down-left / down-right

Distance Levels:

Level	Think of it as...	Physical Displacement
large	half the screen away	20mm
medium	a few icons away	8mm
small	one icon away	3mm
nudge	almost there, fine-tune	1mm

Full Operation Cycle:

1. park()              → Retract stylus out of frame
2. screenshot()        → Clean screenshot, AI sees screen content
3. AI decides          → e.g. "move down-right, large"
4. move(dir, dist)     → Stylus moves toward target
5. screenshot()        → AI checks stylus position (stylus visible)
6. Aligned?
   → No:  back to step 3 (AI re-evaluates and adjusts)
   → Yes: tap()  → Stylus touches screen
7. park()              → Retract stylus out of frame
8. screenshot()        → Verify result, continue next action

Gesture Implementation

Single Tap: G0 to target → stylus down → hold 50-100ms → stylus up

Long Press: G0 to target → stylus down → hold 800ms → stylus up

Swipe: G0 to start → stylus down → G4 P0.03 → G1 to end F3000 → G4 P0.03 → stylus up

Double Tap: stylus down 50ms → up → wait 100ms → stylus down 50ms → up (interval < 300ms)

Use Cases

Scenario	Status
Order food delivery (Meituan, Uber Eats)	Yes (enable password-free or give agent the password for full autonomy)
Hail a ride (Didi, Uber)	Yes (same as above)
Browse and shop (Taobao, Amazon)	Yes (same as above)
Check weather / news / stocks	Fully capable
Read and reply to messages (WeChat, WhatsApp)	Yes
Scroll social media (TikTok, Instagram)	Yes
App daily check-in / collect rewards	Fully capable
Set alarm / timer / reminder	Yes
Take a screenshot and send it	Yes

Security Warning

PhysiClaw has full physical control of your phone — it can see and tap anything on screen. Even without your passwords, it could open your password manager, read saved credentials, receive OTP codes to reset passwords, or access any app that's already logged in. If a malicious actor compromises your agent, they have the same access.

Treat it like handing your unlocked phone to a stranger.

• Use a dedicated backup phone — never your primary device
• Separate phone number — not linked to your main accounts
• Fresh accounts — don't log into your real accounts on it
• Different passwords — never reuse credentials from your primary phone
• Limited funds — only load a small amount of money, enough for the task
• No password manager — don't install one; only store what the agent needs

License

MIT