PIN Client

P2P Inference Network Operator Client - A Tauri-based desktop application that connects your Ollama instance to the PIN network.

Overview

The PIN Client allows GPU/NPU operators to offer inference services through the P2P Inference Network. It:

1. Securely stores credentials in your OS keychain (macOS Keychain, Windows Credential Manager, Linux Secret Service)
2. Maintains a persistent WebSocket connection to the PIN server
3. Proxies inference requests to your local Ollama instance
4. Reports health metrics (load, capacity, models) to the network

Prerequisites

• Ollama installed and running locally
• At least one model pulled (e.g., ollama pull llama3.2)
• PIN operator credentials (client_id and api_secret from registration)

Installation

Pre-built Binaries

Download the latest release for your platform from the Releases page.

Build from Source

1. Install Rust and Tauri prerequisites:

   # Install Rust
   curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh
   
   # macOS
   xcode-select --install
   
   # Ubuntu/Debian
   sudo apt update
   sudo apt install libwebkit2gtk-4.0-dev build-essential curl wget libssl-dev libgtk-3-dev libayatana-appindicator3-dev librsvg2-dev
   sudo apt install -y libwebkit2gtk-4.1-dev libjavascriptcoregtk-4.1-dev libsoup-3.0-dev libgtk-3-dev libayatana-appindicator3-dev librsvg2-dev
   
   # Windows
   # Install Visual Studio Build Tools

2. Clone and build:

   cd pin-client
   cargo tauri build

Configuration

1. Launch the PIN Client

2. Enter your credentials:

   • Client ID: Your pin_xxxx identifier from registration
   • API Secret: Your secret key (stored securely in OS keychain)
   • Ollama URL: Usually http://localhost:11434
   • Server URL: Leave blank for production, or enter custom server

3. Click "Test Ollama" to verify your Ollama connection

4. Click "Save & Connect" to join the network

Architecture

┌─────────────────────────────────────────────────────────────────┐
│                         PIN Client                              │
├─────────────────────────────────────────────────────────────────┤
│  ┌─────────────┐  ┌─────────────┐  ┌─────────────────────────┐ │
│  │  UI (HTML)  │  │  Keychain   │  │    WebSocket Handler    │ │
│  │             │  │  Storage    │  │                         │ │
│  └──────┬──────┘  └──────┬──────┘  └───────────┬─────────────┘ │
│         │                │                     │               │
│         └────────────────┼─────────────────────┤               │
│                          │                     │               │
│  ┌───────────────────────┴─────────────────────┴─────────────┐ │
│  │                    Tauri Commands                          │ │
│  └───────────────────────────────────────────────────────────┘ │
│                          │                                     │
│  ┌───────────────────────┴─────────────────────────────────┐   │
│  │                  Ollama Proxy                            │   │
│  └───────────────────────────────────────────────────────────┘  │
└────────────────────────────────┬───────────────────────────────┘
                                 │
                    ┌────────────┴────────────┐
                    │    Local Ollama         │
                    │    (localhost:11434)    │
                    └─────────────────────────┘

Message Protocol

Authentication

{
  "type": "AUTH",
  "client_id": "pin_xxxx",
  "timestamp": "1704499200",
  "signature": "sha256(client_id + timestamp + secret_hash)"
}

Heartbeat

{ "type": "HEARTBEAT" }
// Response: { "type": "HEARTBEAT_ACK" }

Model List Update

{
  "type": "MODEL_LIST",
  "models": ["llama3.2", "mistral", "codellama"]
}

Inference Request (from server)

{
  "type": "INFERENCE_REQUEST",
  "request_id": "req_abc123",
  "payload": {
    "model": "llama3.2",
    "messages": [{"role": "user", "content": "Hello"}]
  }
}

Inference Response

{
  "type": "INFERENCE_RESPONSE",
  "request_id": "req_abc123",
  "result": {
    "model": "llama3.2",
    "message": {"role": "assistant", "content": "Hi!"},
    "done": true
  }
}

System Tray

The PIN Client runs in your system tray with the following options:

• Status: Shows current connection state
• Connect/Disconnect: Toggle network connection
• Show Window: Open the main window
• Quit: Exit the application

Security

• API secrets are stored in your OS's secure credential storage
• WebSocket connections use TLS encryption
• Authentication uses HMAC-SHA256 signatures with timestamp replay protection

Troubleshooting

"Failed to connect to Ollama"

• Ensure Ollama is running: ollama serve
• Check the URL is correct (default: http://localhost:11434)

"Invalid credentials"

• Verify your client_id and api_secret from the operator dashboard
• Re-register if credentials were regenerated

"Connection timeout"

• Check your internet connection
• Verify the server URL is accessible