# mims-harvard/ToolUniverse

Democratizing AI scientists with ToolUniverse

Repository: https://github.com/mims-harvard/ToolUniverse
Canonical: https://ross.abutalabs.com/products/tooluniverse
Homepage: https://aiscientist.tools
Language: Python
License: Apache-2.0
License Family: permissive
Topics: agents, reasoning-agent, reasoning-language-models, tool-use, ai-agents, ai-scientists, automated-science, co-pilot, co-scientist, ai-for-science, autonomous-agents, llms, lrm, ai-communication, mcp-servers, scientific-skill
Last push: 2026-08-25T11:51:31+00:00

## Health v2 (maintenance only)
Score: 82/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 99, release rhythm 85, longevity 39
- inputs: {"age_days": 548, "days_push": 8, "days_rel": 21, "gap_med": 2.0, "n_releases_24m": 63}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1656, forks 252 (observed 2026-08-28T04:05:18.043623+00:00)

## What it is
ToolUniverse is a Python SDK and MCP server that aggregates a large collection of scientific tools (APIs, databases, local ML and cheminformatics utilities) for use by AI agents. It aims to democratize AI scientists by letting LLM agents call scientific tools through a unified interface.

## Use cases
- give my AI agent access to scientific databases and APIs
- build an autonomous AI scientist that can run bioinformatics tools
- connect scientific tools to Claude via MCP
- let an LLM agent query protein and drug databases
- automate scientific workflows with tool-calling agents
- install agent skills for scientific research tasks

## When to choose
- you are building LLM agents that need to call scientific or bioinformatics tools
- you want an MCP server exposing many scientific APIs with one config
- you are prototyping AI-for-science or automated research pipelines

## When to avoid
- you need a general-purpose agent framework unrelated to science
- you want a polished end-user application rather than an SDK/tool collection
- your tools are not covered and you cannot wrap them yourself

## Facets
- artifact type: library
- maturity: active
- function: agent-framework, mcp, sdk, llm-inference, workflow-automation
- domain: artificial-intelligence, large-language-models, data-science, developer-tools
- platform: python, cross-platform, cli
- tags: ai-scientists, tool-use, mcp-server, scientific-tools, agentic-ai, automated-science, ai-agents

## Member repositories
- mims-harvard/ToolUniverse (main) score 82

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:18.043623+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T03:44:50.088319+00:00, confidence not recorded.
  - readme: https://github.com/mims-harvard/ToolUniverse (fetched 2026-08-28T04:05:18.043623+00:00, sha dff35e01de9c)
  - homepage: https://aiscientist.tools (fetched 2026-08-29T11:17:36.673262+00:00, sha cf597131bd7c)
  - registry_pypi: https://pypi.org/pypi/tooluniverse/json (fetched 2026-08-29T11:17:36.682160+00:00, sha 75a35707fe34)
- Data as of 2026-08-30T08:39:29.467469+00:00.