Start

START: Scalable, Tailored Active-inference Research & Training

Python 3.10+ License: CC BY-SA 4.0 Code style: black

An advanced AI-powered system for creating personalized Active Inference and Free Energy Principle curricula Quick links: activeinference.institute • activities.activeinference.institute • x.com/InferenceActive • discord.activeinference.institute • donate.activeinference.institute • youtube.com/c/ActiveInference • video.activeinference.institute

START combines real-time research capabilities with sophisticated content generation to produce professional-grade, personalized educational materials for Active Inference and the Free Energy Principle. The system integrates multiple APIs, comprehensive prompt engineering, and multilingual capabilities to create world-class curricula tailored to specific domains and individual learners.

📌 Start here: here.md — interactive landing for the full experience. Also see the Docs Hub and this README for GitHub-oriented navigation.

🌐 Documentation (Live)

🔗 Quick Links

▶️ Run the Experience

./run.sh

# Serve locally with live reload
./run_docs.sh --serve

# Build static site to ./site and open it
./run_docs.sh --build

# Deploy to GitHub Pages and open the URL
./run_docs.sh --deploy

📚 Documentation Overview

🚀 Key Features

graph LR
  A[START] --> R[Research]
  A --> C[Content Generation]
  A --> V[Visualizations]
  A --> T[Translation]
  A --> Q[Quality & Testing]
  R --> R1[Perplexity API]
  C --> C1[OpenRouter LLMs]
  V --> V1[Charts & Mermaid]
  T --> T1[11+ Languages]
  Q --> Q1[pytest, ruff, black]

  click R "docs/pipeline.md" "Pipeline"
  click Q "docs/TESTING.md" "Testing"
  click V "docs/pipeline.md" "Visualizations"
  click T "docs/pipeline.md" "Translation"

🔍 Intelligent Research

✍️ Advanced Content Generation

📊 Rich Visualizations

🌍 Multilingual Excellence

🧪 Comprehensive Testing & Quality Assurance

📦 Core Pipeline Scripts

Configuration-Based Research

See docs/getting_started.md for full command lists and script paths.

Learn More

🛠️ Quick Installation

Prerequisites

Installation Steps

# Install uv package manager
curl -LsSf https://astral.sh/uv/install.sh | sh

# Clone and set up project
git clone https://github.com/ActiveInferenceInstitute/Start.git
cd Start

# Install dependencies
uv sync --all-extras --dev

# Download language models
uv run python -m spacy download en_core_web_sm

# Configure API keys
cp .env.example .env
$EDITOR .env  # Add PERPLEXITY_API_KEY and OPENROUTER_API_KEY

# Verify installation
uv run pytest -q
uv run ruff check .
uv run black --check .

🎯 Getting Started

See docs/getting_started.md for first-run commands, generation steps, and exploring outputs.

⚙️ Configuration System

See docs/configuration.md for YAML examples and CLI usage.

📈 Generated Content Quality

Research Analysis (Enhanced with new prompts)

Curriculum Content (Professional-grade)

Multilingual Adaptations (Cultural excellence)

🏗️ System Architecture

flowchart TD
  A[data/config/*] --> B[Research]
  B --> C[Curriculum]
  C --> D[Visualizations]
  D --> E[Translations]
  E --> F[data/* outputs]

  click B "docs/pipeline.md" "Pipeline"
  click F "docs/README.md" "Docs Hub"

See docs/pipeline.md for architecture, templates, and data flow.

Prompt Template System

data/prompts/
├── research_domain_analysis.md     # 6-section domain framework (3K-5K words)
├── research_domain_curriculum.md   # 9-section curriculum generation (40-60 hours)
├── research_entity.md              # 6-section personalization (5K-8K words)
├── curriculum_section.md           # Comprehensive module creation (3-5 hours)
└── translation.md                  # 7-section multilingual framework

Quality Assurance Framework

📊 Example Outputs

See docs/examples.md for example outputs and paths.

🎯 Use Cases & Applications

Educational Institutions

Individual Learning

Organizational Training

📚 Comprehensive Documentation

🚀 Getting Started Guides

🔧 Technical References

📋 Advanced Topics

🤝 Active Inference Ecosystem Integration

See also: implementation repos and knowledge resources in docs/clones.md.

🔄 Development Roadmap

Immediate Enhancements

Future Vision

🤝 Contributing

We welcome contributions! See our Contributing Guide for details on:

Development Workflow

# Set up development environment
uv sync --all-extras --dev

# Run quality checks
uv run pytest -q           # Test suite
uv run ruff check .        # Linting
uv run black --check .     # Formatting

# Development with proper Python path
export PYTHONPATH=$(pwd):$PYTHONPATH

🧪 Development & Testing

Testing Framework

The project includes a comprehensive testing framework with 375+ test cases covering:

Running Tests

# Run full test suite
uv run pytest

# Run specific test categories  
uv run pytest -m "not integration"     # Skip integration tests
uv run pytest -m integration           # Only integration tests
uv run pytest tests/test_domain.py     # Specific test file

# Run with coverage
uv run pytest --cov=src --cov-report=html

# Set environment for GUI-free testing
export MPLBACKEND=Agg
uv run pytest

Development Guidelines

Project Structure

├── src/                    # Core system modules
│   ├── common/            # Shared utilities
│   ├── perplexity/        # API integrations  
│   ├── system/            # System utilities
│   └── terminal/          # CLI components
├── learning/              # Educational pipeline scripts
├── tests/                 # Test suite (375+ tests)
├── docs/                  # Documentation
│   ├── TESTING.md         # Testing guide
│   └── environment.md     # Setup instructions
└── data/                  # Generated content storage

For detailed testing information, see docs/TESTING.md.

📄 License & Citation

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License - see the LICENSE file for details.

DOI & Archive

DOI

This repository is archived and citable via Zenodo: 10.5281/zenodo.17047619

Citation

If you use START in academic work, please cite:

Daniel Ari Friedman, & Active Inference Institute. (2025). 
ActiveInferenceInstitute/Start: v1 (v1). Zenodo. 
https://doi.org/10.5281/zenodo.17047617

START: Scalable, Tailored Active-inference Research & Training
Active Inference Institute (2024)
https://github.com/ActiveInferenceInstitute/Start
Licensed under Creative Commons Attribution-ShareAlike 4.0 International

🙏 Acknowledgments

📬 Contact & Support

🌟 Join the Community

Together we’re building tools to make Active Inference accessible and adaptable across domains, languages, and perspectives. START represents a new paradigm in AI-powered educational content creation - join us in making advanced neuroscience and cognitive science accessible to learners worldwide!

**📚 Explore Documentation 🚀 Get Started 🤝 Join Community**
This site is open source. Improve this page.