Open Science Assistant (OSA)¶

An extensible AI assistant platform for researchers working with open science tools. Built with LangGraph/LangChain and FastAPI.

Overview¶

OSA provides domain-specific AI assistants for open science tools (HED, BIDS, EEGLAB) with:

• Modular tool system for document retrieval, validation, and code execution
• Multi-source knowledge bases from GitHub, OpenALEX, Discourse forums, mailing lists
• Extensible architecture for adding new assistants and tools
• Production-ready observability via LangFuse

Design Philosophy¶

• Precision over features: Researchers need accurate, citation-backed answers
• Simple infrastructure: Lab server deployment, no complex scaling
• Extensible tools: General tool system that communities can adapt
• Domain expertise: Deep knowledge of specific tools, not broad generalist

Target Users¶

• Researchers learning HED annotations, BIDS formatting, or EEGLAB analysis
• Lab members needing quick, accurate answers from documentation
• Developers integrating these tools who need API/usage guidance

Quick Start¶

# Clone the repository
git clone https://github.com/OpenScience-Collective/osa
cd osa

# Install dependencies
uv sync

# Run development server
uv run uvicorn src.api.main:app --reload --port 38528

# CLI usage
uv run osa --help

# Interactive chat with HED assistant
uv run osa chat -a hed

# Single query to HED assistant
uv run osa ask -a hed "How do I annotate a button press?"

Specialist Assistants¶

Documentation¶

• Getting Started - Installation and setup
• Architecture - System design and diagrams
• CLI Reference - Command-line interface
• API Reference - REST API documentation
• Tools - Available tools
• Development - Contributing to OSA

External API Integrations¶

OSA integrates with existing validator and tool APIs rather than hosting validation engines locally:

License¶

MIT