{ "name": "jarvis", "version": "0.1.22", "description": "", "type": "module", "scripts": { "dev": "vite", "build": "tsc && vite build", "lint": "eslint . --ext ts,tsx --report-unused-disable-directives --max-warnings 0", "preview": "vite preview", "format": "prettier --write .", "prepare": "husky", "prod": "node scripts/serverProd.mjs" }, "keywords": [], "author": "Nico Martin mail@nico.dev", "license": "MIT", "dependencies": { "@google/generative-ai": "^0.24.1", "@headlessui/react": "^2.2.5", "@heroicons/react": "^2.2.0", "@huggingface/transformers": "^4.1.0", "@modelcontextprotocol/sdk": "^1.17.0", "@tailwindcss/vite": "^4.1.11", "phonemizer": "^1.2.1", "preact": "^10.25.4", "preact-router": "^4.1.2", "react-hot-toast": "^2.5.2", "strict-url-sanitise": "^0.0.1", "uuid": "^11.1.0", "vad-recorder": "^0.1.1", "vite-tsconfig-paths": "^5.1.4" }, "devDependencies": { "@eslint/js": "^9.31.0", "@preact/preset-vite": "^2.9.4", "@tailwindcss/postcss": "^4.1.11", "@trivago/prettier-plugin-sort-imports": "^5.2.2", "@types/dom-chromium-ai": "^0.0.9", "@types/node": "^24.0.14", "@webgpu/types": "^0.1.64", "autoprefixer": "^10.4.21", "dotenv": "^17.2.0", "eslint": "^9.31.0", "eslint-plugin-react-hooks": "^5.2.0", "eslint-plugin-react-refresh": "^0.4.20", "globals": "^16.3.0", "husky": "^9.1.7", "postcss": "^8.5.6", "prettier": "^3.6.2", "prettier-plugin-tailwindcss": "^0.6.14", "tailwindcss": "^4.1.11", "typescript": "~5.8.3", "typescript-eslint": "^8.37.0", "vite": "^7.0.5", "vite-plugin-html-config": "^2.0.2" }, "prettier": { "plugins": [ "@trivago/prettier-plugin-sort-imports", "prettier-plugin-tailwindcss" ], "singleQuote": false, "trailingComma": "es5", "importOrder": [ "<THIRD_PARTY_MODULES>", "^[./]" ], "importOrderSeparation": true, "importOrderSortSpecifiers": true } }

JARVIS

A modern AI assistant web application featuring voice interaction, real-time conversation, and MCP (Model Context Protocol) server integration.

Features

• Voice Interaction: Voice activity detection, speech-to-text, and text-to-speech capabilities
• Real-time Chat: Interactive conversation interface with AI assistant
• MCP Server Integration: Connect and interact with Model Context Protocol servers
• Image-to-Text: Convert images to text descriptions
• Multi-modal Interface: Switch between voice-activated JARVIS mode and traditional chat
• Browser-based AI: Runs locally in your browser using WebGPU and language model APIs

Tech Stack

• Framework: Preact with TypeScript
• Styling: Tailwind CSS
• AI/ML: Transformers.js
• Build Tool: Vite
• MCP: Model Context Protocol SDK for server integrations

Getting Started

Prerequisites

• Node.js (version 18 or higher)
• Modern browser with WebGPU support

Installation

1. Clone the repository:

git clone <repository-url>
cd jarvis

2. Install dependencies:

npm install

3. Start the development server:

npm run dev

4. Open your browser and navigate to http://localhost:8080

Scripts

• npm run dev - Start development server
• npm run build - Build for production
• npm run lint - Run ESLint
• npm run format - Format code with Prettier
• npm run preview - Preview production build

Usage

JARVIS Mode

• Click to activate voice interaction
• Speak naturally to interact with the AI assistant
• Visual rings indicate voice activity and responses

Chat Mode

• Switch to traditional text-based chat interface
• Type messages and receive AI responses

MCP Settings

• Navigate to /mcp to configure Model Context Protocol servers
• Add HTTP servers for extended functionality
• Manage server connections and available tools

License

MIT License - see LICENSE file for details.

Author

Nicolas Martin (mail@nico.dev)