tooluniverse

Routes scientific research queries to specialized ToolUniverse skills or general multi-database tool strategies.

Updated Mar 13, 2026
One-click install
npx skills add https://github.com/matt-grain/pharma-catalyst --skill tooluniverse-matt-grain
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: tooluniverse
Source: https://github.com/matt-grain/pharma-catalyst/tree/main/.claude/skills/tooluniverse
Command: npx skills add https://github.com/matt-grain/pharma-catalyst --skill tooluniverse-matt-grain

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Scientific research questions span dozens of domains—drug discovery, genomics, clinical trials, disease profiling—and choosing the right tools and databases for each is overwhelming. This Skill acts as an intelligent router that matches your question to one of 58+ specialized research skills, or falls back to proven general strategies for querying 1400+ scientific tools. ## Core Features & Use Cases - Keyword-Based Routing: Maps user questions to specialized skills covering disease research, drug profiling, variant interpretation, GWAS analysis, clinical trials, pharmacovigilance, and more. - General Research Strategies: When no specialized skill matches, provides workflows for tool discovery, multi-hop query chains, cross-database verification, disambiguation, and evidence-graded report generation. - Fallback Chains: Documents primary-to-fallback tool mappings (e.g., PubMed → EuropePMC → OpenAlex) so research continues even when tools fail. - Use Case: A user asks "What drugs target TP53?" The router matches keywords to the target-research skill, which then chains OpenTargets, ChEMBL, and DGIdb queries to produce a cited, evidence-graded report. ## Quick Start Ask a scientific research question such as "research the safety profile of metformin" and the router will invoke the appropriate specialized skill automatically.

Frequently Asked Questions about tooluniverse

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I research a drug, disease, or gene with ToolUniverse?▼

Ask your research question in natural language, such as "profile metformin" or "research Alzheimer's disease." The router extracts keywords and invokes the matching specialized skill, which chains database queries and produces a cited report.

What databases does ToolUniverse cover for biomedical research?▼

ToolUniverse tools span PubMed, UniProt, ChEMBL, PubChem, OpenTargets, ClinVar, gnomAD, GTEx, Reactome, ClinicalTrials.gov, FAERS, and more. Each data type has primary and fallback databases for cross-referencing.

What happens when no specialized skill matches my question?▼

The router falls back to general strategies: exhaustive tool discovery via find_tools queries, multi-hop tool chains, multi-database querying, and disambiguation. It executes actual queries rather than just describing steps.

How does ToolUniverse handle tool failures during research?▼

Failed tools are retried, then replaced with documented fallbacks—for example, PubMed falls back to EuropePMC then OpenAlex. If all options fail, the gap is documented as unavailable rather than left blank.

Can I use ToolUniverse in languages other than English?▼

Yes. You can write questions in any language and receive answers in that language. Internally, all tool queries are translated to English since the underlying scientific databases operate in English.

What are the limitations of the ToolUniverse router?▼

The router depends on keyword matching, so ambiguous questions like "tell me about aspirin" trigger a clarification prompt. Some tools require API keys configured via environment variables, and unavailable data is documented rather than estimated.