ToolUniverse
About this tool
An AI-scientist tool ecosystem from Harvard Medical School's Zitnik Lab. It exposes more than 1,000 ML models, datasets, APIs and scientific packages through a common interface. Areas include drug discovery, precision oncology, rare disease, pharmacovigilance, chemistry, bioinformatics and literature search. Connect through native MCP, agent skills or a Claude Code plug-in.
- Feature 1: Start the native MCP server with uvx tooluniverse.
- Feature 2: More than 1,000 models, datasets, APIs and scientific packages.
- Feature 3: Domain skills for drug discovery, precision oncology, rare disease and pharmacovigilance.
- Feature 4: Installation through npx skills add or the Claude Code plug-in.
- Feature 5: The tu CLI discovers, runs and tests tools.
Resource information
- Maintainer
- Zitnik Lab, Harvard Medical School; led by Shanghua Gao and Marinka Zitnik.
- Type
- MCP
- Category
- Discipline-specific tools
- Runs on
- MCP
- Who it's for
- Biomedical, pharmaceutical, chemistry and bioinformatics researchers querying sources such as FDA, PubChem and PubMed, exploring targets and drug safety, and combining clinical data.
- License at time of listing
- Apache-2.0
- Source
- GitHub
