Skip to main content
ClaudeWave
Skill1.4k repo starsupdated today

tooluniverse-claude-code-plugin

The ToolUniverse Claude Code plugin installs a managed MCP server providing 1000+ scientific tools, 120+ research skills, and automated routing in a single command. Use this plugin for one-step setup, troubleshooting plugin loading issues, verifying MCP server connectivity, checking API key requirements, or enabling auto-updates for ToolUniverse capabilities.

Install in Claude Code
Copy
git clone --depth 1 https://github.com/mims-harvard/ToolUniverse /tmp/tooluniverse-claude-code-plugin && cp -r /tmp/tooluniverse-claude-code-plugin/plugin/skills/tooluniverse-claude-code-plugin ~/.claude/skills/tooluniverse-claude-code-plugin
Then start a new Claude Code session; the skill loads automatically.

SKILL.md

# Install the ToolUniverse Plugin for Claude Code

One-step install of the ToolUniverse plugin: MCP server with 1000+ tools, 120+ research skills, slash commands, and the research agent — all auto-configured.

## Prerequisites (check once)

```bash
uv --version      # must exist; if not:  curl -LsSf https://astral.sh/uv/install.sh | sh
claude --version  # Claude Code CLI; if not: https://claude.com/claude-code
```

## Install (two commands)

```bash
# 1. Register the ToolUniverse marketplace from GitHub
claude plugin marketplace add mims-harvard/ToolUniverse

# 2. Install the plugin
claude plugin install tooluniverse@tooluniverse
```

That's it. Restart Claude Code. The MCP server auto-starts via `uvx tooluniverse` on first use (~30 s cold start, instant after).

### Important: Remove global skills if previously installed

If you previously installed ToolUniverse skills globally (via `tooluniverse-install-skills` or manual copy), **remove them**. The plugin includes all skills — global copies interfere with the plugin's skill routing.

```bash
# Check for global skills
ls ~/.claude/skills/tooluniverse-* 2>/dev/null | wc -l

# Remove them (the plugin replaces them)
rm -rf ~/.claude/skills/tooluniverse-*
rm -rf ~/.claude/skills/create-tooluniverse-skill
rm -rf ~/.claude/skills/setup-tooluniverse
```

### Pin the tool version (optional, for reproducibility)

The MCP server runs the `tooluniverse` PyPI package via `uvx`. To lock tool behavior for a long-running analysis, pin the package version in the plugin's `.mcp.json` — change the args to:

```json
"args": ["tooluniverse@1.2.2"]
```

Replace `1.2.2` with any released version from https://pypi.org/project/tooluniverse/. Restart Claude Code to apply. (`.mcp.json` location is shown under "API keys" below.)

## Verify it worked

```bash
claude plugin list
# Expect: tooluniverse  (enabled)
```

Inside Claude Code, just ask naturally:
```
What are the top mutated genes in breast cancer?
Research the drug metformin.
```

The router skill auto-dispatches to the right specialized skill — no command prefix needed.

## What you get

| Component | What it does | How to invoke |
|---|---|---|
| **MCP server** | 1000+ tools via `find_tools`, `get_tool_info`, `execute_tool` | Auto-loaded; no action needed |
| **`/tooluniverse:research`** | Multi-database investigation inline in this chat (you see each step) | Slash command |
| **`/tooluniverse:translate-id`** | Resolve an ID across all relevant namespaces | Slash command |
| **`/tooluniverse:cross-validate`** | Verify a claim across 3+ independent databases | Slash command |
| **`/tooluniverse:compare`** | N-way side-by-side comparison with domain-appropriate columns | Slash command |
| **`/tooluniverse:literature-sweep`** | Graded mini-review across PubMed + EuropePMC + Semantic Scholar | Slash command |
| **`/tooluniverse:researcher`** | Same investigation as `research`, delegated to a forked subagent | Slash command |
| **120+ skills** | Structured workflows (drug research, variant interpretation, pharmacovigilance, CRISPR screens, statistical modeling, etc.) | Auto-activate on matching questions |

## API keys (optional, but recommended)

Most tools work without keys. For enhanced access, add the keys you care about to the MCP server's `env` block. Locate the installed plugin's `.mcp.json`:

```bash
# Show component inventory + on-disk location:
claude plugin details tooluniverse

# Or find the file directly:
find ~/.claude/plugins -name '.mcp.json' -path '*tooluniverse*'

# Typical install location after marketplace install:
# ~/.claude/plugins/cache/tooluniverse/tooluniverse/<version>/.mcp.json
```

```json
{
  "mcpServers": {
    "tooluniverse": {
      "command": "uvx",
      "args": ["tooluniverse"],
      "env": {
        "PYTHONIOENCODING": "utf-8",
        "NCBI_API_KEY": "your_key",
        "NVIDIA_API_KEY": "your_key",
        "ONCOKB_API_TOKEN": "your_token"
      }
    }
  }
}
```

Full API-key list: `setup-tooluniverse` skill → `API_KEYS_REFERENCE.md`.

## Update

```bash
claude plugin update tooluniverse
# Also refresh the MCP server's tool cache:
uv cache clean tooluniverse
```

Restart Claude Code.

## Uninstall

```bash
claude plugin uninstall tooluniverse
claude plugin marketplace remove tooluniverse
```

## Alternative install paths

### Clone + local install (air-gapped or forks)

```bash
git clone https://github.com/mims-harvard/ToolUniverse.git
cd ToolUniverse
claude plugin marketplace add ./
claude plugin install tooluniverse@tooluniverse
```

### Download the release zip (no git needed)

Grab `tooluniverse-plugin-vX.Y.Z.zip` from https://github.com/mims-harvard/ToolUniverse/releases, unzip, then:

```bash
claude plugin marketplace add /path/to/unzipped-dir
claude plugin install tooluniverse@tooluniverse
```

## Troubleshooting

| Symptom | Fix |
|---|---|
| `marketplace add` fails with "no marketplace.json" | Ensure you're using `mims-harvard/ToolUniverse` (owner/repo), not the plugin subdir. |
| `uvx: command not found` | Install `uv` (see Prerequisites), reopen terminal. |
| MCP server won't start | Test in terminal: `uvx tooluniverse`. If it fails there, it's an `uv`/Python issue. |
| Plugin installs but tools missing | Restart Claude Code. First launch downloads the package (~30 s). |
| `requires-python >= 3.10` | `uv python install 3.12` |
| Tools feel outdated | `uv cache clean tooluniverse` then restart Claude Code. |
| New release not picked up / tools run an OLD version | Run `uv tool list \| grep tooluniverse`. If it's listed, a global `uv tool install` is **shadowing** the plugin's `uvx` (uvx prefers an installed tool over the latest published version). Fix: `uv tool uninstall tooluniverse`, then restart. Normal users without that install always get the latest automatically. |

Still stuck: https://github.com/mims-harvard/ToolUniverse/issues

## Automated diagnosis & repair

If the table above didn't fix it, hand this skill to an agent — *"read the
tooluniverse-c
setup-tooluniverseSkill

Install and configure ToolUniverse for any use case — MCP server (chat-based), CLI (command line with 9 subcommands), or Python SDK (Coding API with 3 calling patterns). Covers uv/uvx setup, MCP configuration for 12+ AI clients (Cursor, Claude Desktop, Windsurf, VS Code, Codex, Gemini CLI, Trae, Cline, etc.), full CLI reference (tu list/grep/find/info/run/test/status/build/serve), Coding API quickstart, agentic tools, code executor, API key walkthrough, skill installation, and upgrading. Use when user asks how to set up ToolUniverse, which access mode to use (MCP vs CLI vs SDK), configuring MCP servers, using the CLI, troubleshooting installation, upgrading, or mentions installing ToolUniverse or setting up scientific tools. Also triggers for "how do I use ToolUniverse", "what's the best way to access tools", "command line", "tu command", "coding API", "tu build".

tooluniverse-acmg-variant-classificationSkill

Systematic ACMG/AMP germline variant classification with all 28 criteria (PVS1, PS1-4, PM1-6, PP1-5, BA1, BS1-4, BP1-7) for clinical significance. Produces 5-tier verdict (Pathogenic / Likely Pathogenic / VUS / Likely Benign / Benign) with cited evidence per criterion. Use for variant interpretation, VUS resolution, and pathogenicity assessment. Combines ClinVar, gnomAD, computational predictors, and gene-mechanism context.

tooluniverse-admet-predictionSkill

Comprehensive ADMET (Absorption, Distribution, Metabolism, Excretion, Toxicity) profiling for drug candidates. Integrates ADMET-AI predictions, SwissADME drug-likeness, PubChemTox experimental toxicity, ChEMBL clinical data, Lipinski rule-of-five, and CYP interaction data. Use for drug-likeness assessment, BBB penetration, bioavailability, hepatotoxicity prediction, ADME/PK profiling, or screening compound libraries before lab testing.

tooluniverse-adverse-event-detectionSkill

Detect and analyze adverse drug event signals using FDA FAERS reports, drug labels, and disproportionality statistics (PRR, ROR, IC). Generates quantitative safety signal scores (0-100) with evidence grading. Use for post-market surveillance, pharmacovigilance, drug safety assessment, regulatory submissions, and detecting rare AE signals not visible in clinical trials.

tooluniverse-adverse-outcome-pathwaySkill

Map environmental and industrial chemicals to adverse outcome pathways (AOPs) — molecular initiating event to organ-level toxicity. Uses AOPWiki, GHS classification, IARC carcinogen status, and LD50 data. Use for environmental/industrial chemical risk assessment, regulatory-grade hazard characterization, and AOP stressor mapping. Distinct from drug-safety analysis (use tooluniverse-pharmacovigilance for drugs).

tooluniverse-aging-senescenceSkill

Aging biology, cellular senescence, and longevity research. Covers senescence markers (p16/CDKN2A, SASP, SA-beta-gal), aging hallmarks, senolytic drug discovery (dasatinib+quercetin, fisetin, navitoclax), epigenetic clocks, telomere biology, and longevity GWAS. Use for senescence-pathway analysis, age-related disease genetics, senolytic-target discovery, and centenarian-genetics queries. Distinguishes correlative vs causal evidence (knockout, intervention).

tooluniverse-antibody-engineeringSkill

Therapeutic antibody engineering and optimization, lead-to-clinical-candidate. Covers sequence humanization (germline alignment, framework retention), affinity maturation, developability (aggregation, stability, PTMs), structure modeling (AlphaFold/PDB CDR analysis), immunogenicity prediction, and manufacturing feasibility. Use for biologic-drug optimization, mAb design review, biosimilar engineering, and clinical-precedent comparison.

tooluniverse-binder-discoverySkill

Discover novel small-molecule binders for protein targets using structure-based and ligand-based screening. Covers druggability assessment, known-ligand mining (ChEMBL, BindingDB), similarity expansion, ADMET filtering, and synthesis feasibility. Use for hit identification, virtual screening, target-to-compounds workflows, and lead-finding before commit-to-medchem.