Chemical Safety (ToolUniverse Claude Skill)

A ToolUniverse agent skill that assesses a compound’s toxicological risk by combining predictive toxicology, toxicogenomics, structural alerts, and regulatory safety data into a graded risk profile.

   
Type Claude Skill (one of ToolUniverse’s pre-built agent skills)
Supplier Zitnik Lab, Harvard Medical School
Availability GA — part of the ToolUniverse skills collection (skills/tooluniverse-chemical-safety/)
Pricing Free / OSS (Apache-2.0); wraps public APIs (ADMET-AI, CTD, PubChem/PubChemTox, AOPWiki, ChEMBL, STITCH, FDA, DrugBank)
Capabilities Read-only — drives ToolUniverse tool calls; no data writes
Verified works · 2026-07-20
Security cleared · 2026-07-20 — provenance matches Zitnik Lab, Apache-2.0, skill dir confirmed, no OSV advisories

How to install

This skill calls ToolUniverse tools, so the ToolUniverse MCP server must be installed first (see the ToolUniverse page). Simplest registration:

claude mcp add --transport stdio tooluniverse -- uvx tooluniverse

Then add the skills:

  • Claude Code — install the whole skill collection (the skill resolves as tooluniverse-chemical-safety): 
    npx skills add mims-harvard/ToolUniverse
  • Manual / other agents — copy just this skill directory into your skills folder: 
    git clone https://github.com/mims-harvard/ToolUniverse
cp -r ToolUniverse/skills/tooluniverse-chemical-safety ~/.claude/skills/

    (replace ~/.claude/skills/ with your agent’s skills directory if you are not using Claude Code/Desktop.)

The skill sets disable-model-invocation: true upstream, so invoke it explicitly (e.g. ask Claude to “use the chemical-safety skill”) rather than relying on automatic dispatch.

What it does

Runs an eight-phase toxicology pipeline:

  1. Compound disambiguation — resolve identity to SMILES, PubChem CID, ChEMBL ID.
  2. Predictive toxicology — ADMET-AI endpoints (AMES mutagenicity, DILI, LD50, carcinogenicity, hERG).
  3. ADMET properties — absorption, distribution, metabolism, excretion, CYP interactions.
  4. Toxicogenomics — CTD chemical-gene-disease mapping; PubChemTox experimental data; AOPWiki adverse-outcome pathways.
  5. Regulatory safety — FDA boxed warnings, contraindications, adverse reactions (pharmaceuticals only).
  6. Drug safety profile — DrugBank toxicity, contraindications, interactions (pharmaceuticals only).
  7. Chemical-protein interactions — STITCH binding networks, off-target effects.
  8. Structural alerts — ChEMBL PAINS/Brenk/Glaxo alerts.

It synthesizes an integrated risk classification (Critical/High/Medium/Low) with T1–T4 evidence grading.

Primary use cases: hazard identification, occupational/consumer-product toxicity screening, dose-response evaluation, acute vs. chronic toxicity assessment, and safety triage of drug candidates.

Notes

It is a reasoning layer over ToolUniverse; without the MCP server registered, the tool calls fail. Distinct from the Small Molecule Discovery skill (identity/activity/sourcing) and the Adverse Outcome Pathway and Adverse Event Detection skills — this one is a compound-level toxicology profile. ToolUniverse ships ~68 such skills; other workflows are catalogued separately.

Sources

Installed this tool?

Share feedback — install path, OS, errors, workarounds. The form opens with this tool pre-selected and a link back to this page.