🧪 Autonomous AI Researcher

An advanced, agentic AI platform designed to automate scientific literature review and research hypothesis generation. This system acts as an autonomous researcher that can search through millions of scientific papers, synthesize information, and brainstorm novel ideas.

🚀 Vision

This repository is evolving into a fully Autonomous AI Researcher. The goal is to create an agent that doesn't just answer questions, but proactively explores scientific frontiers, validates claims against the literature, and proposes high-impact research directions.

✨ Key Features

• 🔍 Intelligent Literature Search: Integrated with Semantic Scholar API to find the most relevant and high-impact papers.
• 🧠 Automated Brainstorming: Generates novel research hypotheses based on real-world context and found literature.
• 🛡️ Human-in-the-Loop (HITL): Advanced Approve/Decline gate for tool execution, giving you full control over the researcher's actions.
• 🌓 Premium Research UI: A state-of-the-art dark theme with glassmorphism, mathematical notation support (KaTeX), and syntax highlighting.
• ⚡ Multi-Model Support: Switch seamlessly between OpenRouter (Nemotron, Mistral) and Google Gemini 2.0.

🛠️ Architecture

• Framework: Next.js 15 (App Router)
• Agent Orchestration: Custom SSE (Server-Sent Events) streaming with tool-calling capabilities.
• Styling: Tailwind CSS with Framer Motion animations.
• APIs: Semantic Scholar Graph API for research data.

🚦 Getting Started

1. Prerequisites

You will need API keys for:

• OpenRouter
• Google AI Studio (for Gemini)
• Semantic Scholar (Optional, public searches work without it)

2. Environment Setup

Create a .env.local file in the root directory:

OPENROUTER_API_KEY=your_key_here
GEMINI_API_KEY=your_key_here
NEXT_PUBLIC_MOCK_LITERATURE=false

3. Installation

npm install
npm run dev

📈 Roadmap

• Basic Chat Interface
• Tool Calling Integration (Search + Brainstorm)
• Human-in-the-Loop Approval Gate
• Scientific Math (LaTeX) Rendering
• Autonomous Report Generation (PDF Export)
• Direct ArXiv Integration for latest pre-prints
• Multi-agent collaboration (Reviewer vs. Researcher)

🤝 Contributing

Contributions are welcome! If you're passionate about agentic scientific workflows, feel free to open a PR or Issue.

---

Built with ❤️ for the future of science.