Aura — AI Search Assistant

Aura is a blazing fast, privacy-focused search engine with an integrated AI assistant. It compiles to 100% native Rust using Axum on the backend and Leptos (WebAssembly) on the frontend. It proxies queries through SearXNG to protect privacy, and processes AI queries using a user-configured self-hosted Ollama server.

---

🏛️ Architecture & Stack

• Frontend: Leptos (WASM)
• Backend: Axum (Rust) / Tokio / SearXNG / Ollama
• Deployment: Nix-built Container / Unraid native / Docker Compose

---

🟢 Key Features

• Privacy-First Search: Proxies queries through SearXNG to ensure absolute user anonymity.
• Integrated AI Assistant: Conversational answers using a local self-hosted Ollama instance.
• Access PIN Security: Lock down the interface with an optional numerical PIN for absolute privacy.
• Performance First: Tiny resource footprint, zero external JS engine dependencies, and rapid page load speeds.

---

💾 Deployment & Installation

Docker Compose

Create a docker-compose.yml file with the following service definition:

services:
  aura:
    image: ubermetroid/aura:latest
    container_name: aura
    restart: unless-stopped
    ports:
      - ${PORT:-4408}:4408
    volumes:
      - ${AURA_DATA_PATH:-./data}:/app/data
    environment:
      PORT: 4408
      SITE_TITLE: ${AURA_SITE_TITLE:-Aura}
      AURA_PIN: ${AURA_PIN:-}
      BASE_URL: ${AURA_BASE_URL:-http://localhost:4408}
      ALLOWED_ORIGINS: ${AURA_ALLOWED_ORIGINS:-*}
      TZ: ${TZ:-UTC}
      ENABLE_TRANSLATION: ${ENABLE_TRANSLATION:-false}
      ENABLE_THEMES: ${ENABLE_THEMES:-true}
      ENABLE_PRINT: ${ENABLE_PRINT:-true}
      MAX_ATTEMPTS: ${MAX_ATTEMPTS:-5}
      OLLAMA_BASE_URL: ${OLLAMA_BASE_URL:-http://localhost:11434}
      OLLAMA_MODEL: ${OLLAMA_MODEL:-llama3}

---

⚙️ Configuration Options

Environment Variable	Description	Default
PORT	The port number the backend HTTP server will bind to inside the container.	4408
SITE_TITLE	Custom website title rendered in navigation headers and browser tabs.	Aura
ALLOWED_ORIGINS	Comma-separated list of allowed HTTP request origins (CORS filter).	*
AURA_PIN	Optional 4–10 digit PIN (numerical only) to lock access to the interface.	None
TZ	Timezone for the container processes and logs.	UTC
ENABLE_TRANSLATION	Enable the multi-language / translation selector in the navigation header.	true
MAX_ATTEMPTS	Number of failed PIN attempts permitted before rate lockout.	5
OLLAMA_BASE_URL	Base URL of the self-hosted Ollama server for text inference.	http://localhost:11434
OLLAMA_MODEL	Model ID to use for chat inference on the Ollama server.	llama3
STATIC_DIR	Directory containing static compiled frontend assets served by Axum.	./frontend/dist

---

🛠️ Local Development

Ensure you have the Rust toolchain installed.

# 1. Run tests
cargo test

# 2. Run clippy checks
cargo clippy --workspace --all-targets

# 3. Start development server
cargo run

---

📄 License

Licensed under the Apache License, Version 2.0. Copyright 2026 UberMetroid.