🚀 DeepCritical¶

Hydra-configured, Pydantic Graph-based deep research workflow

DeepCritical isn't just another research assistant—it's a framework for building entire research ecosystems. While a typical user asks one question, DeepCritical generates datasets of hypotheses, tests them systematically, runs simulations, and produces comprehensive reports—all through configurable Hydra-based workflows.

✨ Key Features¶

• 🔧 Hydra Configuration: Flexible, composable configuration system
• 🔄 Pydantic Graph: Stateful workflow execution with type safety
• 🤖 Multi-Agent System: Specialized agents for different research tasks
• 🧬 PRIME Integration: Protein engineering workflows with 65+ tools
• 🔬 Bioinformatics: Multi-source data fusion and reasoning
• 🌐 DeepSearch: Web research automation
• 📊 Comprehensive Tooling: RAG, analytics, and execution environments

🚀 Quick Start¶

# Install with uv (recommended)
uv sync

# Run a simple research query
uv run deepresearch question="What is machine learning?"

# Enable PRIME flow for protein engineering
uv run deepresearch flows.prime.enabled=true question="Design a therapeutic antibody"

🏗️ Architecture Overview¶

graph TD
    A[Research Question] --> B[Hydra Config]
    B --> C[Pydantic Graph]
    C --> D[Agent Orchestrator]
    D --> E[PRIME Flow]
    D --> F[Bioinformatics Flow]
    D --> G[DeepSearch Flow]
    E --> H[Tool Registry]
    F --> H
    G --> H
    H --> I[Results & Reports]

📚 Documentation¶

• Getting Started - Installation and setup
• Architecture - System design and components
• Flows - Available research workflows
• Tools - Tool ecosystem and registry
• API Reference - Complete API documentation
• Examples - Usage examples and tutorials

🤝 Contributing¶

We welcome contributions! Please see our Contributing Guide for details.

📄 License¶

This project is licensed under the MIT License - see the LICENSE file for details.

📊 Project Status¶