tooluniverse — Claude Code MCP server: what it does, how to install
Connects AI agents to research literature, scientific databases, and computational biology tools. Useful for running automated research workflows in agent sessions.
claude mcp add --env OPENAI_API_KEY=<OPENAI_API_KEY> --env ANTHROPIC_API_KEY=<ANTHROPIC_API_KEY> --env GEMINI_API_KEY=<GEMINI_API_KEY> --env OPENROUTER_API_KEY=<OPENROUTER_API_KEY> --env AZURE_OPENAI_API_KEY=<AZURE_OPENAI_API_KEY> --transport stdio tooluniverse -- uvx tooluniverseWhat tooluniverse does
The server gives coding agents direct access to scientific papers from sources like PubMed and ArXiv. It connects models to specialized biology packages and experimental pipelines. To preserve context space, compact mode swaps the large tool catalog for four or five discovery functions. That keeps the prompt lean.
The system includes a two-tier cache using memory and SQLite storage. Tasks like molecular docking run in the background through async calls. Developers can inspect tools via the tu command-line interface, and tooluniverse-doctor flags missing extra dependencies.
Running the server requires Python 3.10 or later. On Apple Silicon Macs, macOS 14 or newer is necessary because faiss-cpu lacks precompiled arm64 wheels for older releases.
What it helps with
- query biomedical literature across PubMed and ArXiv
- execute protein docking and molecular simulations asynchronously
- reduce context token consumption using compact discovery mode
- diagnose missing package dependencies with
tooluniverse-doctor
Example prompts for the agent
- Search PubMed for recent clinical trials on rare genetic diseases and summarize them.
- Run an ArXiv search for protein folding papers and extract the top citations.
- Check missing extras and tool dependencies using
tooluniverse-doctor.
How to install tooluniverse
Platform CLI
claude mcp add --env OPENAI_API_KEY=<OPENAI_API_KEY> --env ANTHROPIC_API_KEY=<ANTHROPIC_API_KEY> --env GEMINI_API_KEY=<GEMINI_API_KEY> --env OPENROUTER_API_KEY=<OPENROUTER_API_KEY> --env AZURE_OPENAI_API_KEY=<AZURE_OPENAI_API_KEY> --transport stdio tooluniverse -- uvx tooluniverseReplace <…> with your own values. 12 more environment variables are in the server README. uvx runs the command named after the package; if the README names another one, use uvx --from tooluniverse <command>
Platform CLI
codex mcp add tooluniverse --env OPENAI_API_KEY=<OPENAI_API_KEY> --env ANTHROPIC_API_KEY=<ANTHROPIC_API_KEY> --env GEMINI_API_KEY=<GEMINI_API_KEY> --env OPENROUTER_API_KEY=<OPENROUTER_API_KEY> --env AZURE_OPENAI_API_KEY=<AZURE_OPENAI_API_KEY> -- uvx tooluniverseReplace <…> with your own values. 12 more environment variables are in the server README. uvx runs the command named after the package; if the README names another one, use uvx --from tooluniverse <command>
Config file
{
"mcpServers": {
"tooluniverse": {
"command": "uvx",
"args": [
"tooluniverse"
],
"env": {
"OPENAI_API_KEY": "<OPENAI_API_KEY>",
"ANTHROPIC_API_KEY": "<ANTHROPIC_API_KEY>",
"GEMINI_API_KEY": "<GEMINI_API_KEY>",
"OPENROUTER_API_KEY": "<OPENROUTER_API_KEY>",
"AZURE_OPENAI_API_KEY": "<AZURE_OPENAI_API_KEY>"
}
}
}
}Replace <…> with your own values. 12 more environment variables are in the server README. uvx runs the command named after the package; if the README names another one, use uvx --from tooluniverse <command>
Config file
{
"servers": {
"tooluniverse": {
"type": "stdio",
"command": "uvx",
"args": [
"tooluniverse"
],
"env": {
"OPENAI_API_KEY": "<OPENAI_API_KEY>",
"ANTHROPIC_API_KEY": "<ANTHROPIC_API_KEY>",
"GEMINI_API_KEY": "<GEMINI_API_KEY>",
"OPENROUTER_API_KEY": "<OPENROUTER_API_KEY>",
"AZURE_OPENAI_API_KEY": "<AZURE_OPENAI_API_KEY>"
}
}
}
}Replace <…> with your own values. 12 more environment variables are in the server README. uvx runs the command named after the package; if the README names another one, use uvx --from tooluniverse <command>
Platform CLI
gemini mcp add -e OPENAI_API_KEY=<OPENAI_API_KEY> -e ANTHROPIC_API_KEY=<ANTHROPIC_API_KEY> -e GEMINI_API_KEY=<GEMINI_API_KEY> -e OPENROUTER_API_KEY=<OPENROUTER_API_KEY> -e AZURE_OPENAI_API_KEY=<AZURE_OPENAI_API_KEY> tooluniverse uvx tooluniverseReplace <…> with your own values. 12 more environment variables are in the server README. uvx runs the command named after the package; if the README names another one, use uvx --from tooluniverse <command>
Step by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about tooluniverse
How do I install tooluniverse in Claude Code?
Run `claude mcp add --env OPENAI_API_KEY=<OPENAI_API_KEY> --env ANTHROPIC_API_KEY=<ANTHROPIC_API_KEY> --env GEMINI_API_KEY=<GEMINI_API_KEY> --env OPENROUTER_API_KEY=<OPENROUTER_API_KEY> --env AZURE_OPENAI_API_KEY=<AZURE_OPENAI_API_KEY> --transport stdio tooluniverse -- uvx tooluniverse` in your terminal. Replace the placeholders with your actual API keys.
What does tooluniverse do?
It connects AI models to scientific databases, literature repositories, and computational biology packages. The server lets agents retrieve research papers and coordinate multi-step scientific workflows.
What are the system requirements for tooluniverse?
You need Python 3.10 or newer. On Apple Silicon Macs, macOS 14 or higher is required because faiss-cpu does not provide arm64 wheels for earlier OS versions.
