Scimantic

Scimantic is a VS Code extension and semantic framework that enables human-AI teaming for scientific research. It transforms the scientific process into a machine-readable workflow from inception through publication.

Vision

Scimantic transforms the traditional scientific workflow into a fully semantic, machine-readable process by capturing the complete research lifecycle as linked data. By integrating nanopublications, RDF/OWL ontologies, W3C PROV provenance, and uncertainty quantification from the very beginning of research, Scimantic enables:

• Semantic Publishing from Day One: Every research artifact (literature notes, hypotheses, experimental designs, data, analyses) is captured as machine-readable linked data
• Complete Provenance: Full traceability from initial literature review through final publication using W3C PROV-O
• Uncertainty Quantification: Explicit representation of uncertainties in hypotheses, measurements, and conclusions
• AI-Assisted Research: MCP-based AI agents that understand and interact with the semantic graph
• Reproducible Science: Machine-readable workflows that can be validated, reproduced, and extended

The Scientific Method as Linked Data

Scimantic maps the activities in the scientific method to semantic web standards:

1. Question Formation → Generating a Question
2. Literature Search → Extracting Evidence from sources
3. Evidence Assessment → Evaluating Evidence credibility
4. Hypothesis Formation → Deriving a Hypothesis from evidence
5. Design of Experiment → Defining an Experimental Method to test the Hypothesis
6. Experimentation → Executing the Experimental Method to produce a Dataset
7. Analysis → Processing the Dataset to produce Results
8. Result Assessment → Comparing Results with Hypothesis to generate Conclusions

Documentation

For a deeper dive into Scimantic's rationale and design, please refer to the core documentation:

• Vision (Why): The philosophy of semantic research and the reasoning behind Scimantic.
• Architecture (What): High-level system design and component breakdown.
• Roadmap (When): Planned features and milestones.
• Specifications (How): Detailed ontology and technical specifications.
• Features: Vertical slice implementation plans.
• Release Guide: How to release Ontology, Core, and Extension versions.

Development

This repository is a uv workspace.

# Install dependencies for core
uv sync

# Run tests
cd scimantic-core && uv run pytest

# Start MCP server (for VS Code extension)
cd scimantic-core && uv run python -m scimantic.mcp

Quality Assurance & Automation

This project uses pre-commit to ensure code quality, consistency, and documentation accuracy.

To enable the automation:

# Install git hooks
pre-commit install

This includes ontology validation:

• Syntax: Validates scimantic.ttl format
• Visualization: Auto-generates ontology_graph.png (Mermaid) whenever the ontology changes

License

[Add license information]