ToolUniverse is a collection of tools, interfaces, and supporting components for building AI systems that perform scientific work. It is for developers creating AI scientist agents that use APIs, databases, machine-learning tools, and domain-specific utilities. The catalogue includes skills, commands, an MCP server, an agent, and a hook for working with the ecosystem.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add mims-harvard/ToolUniverse --skill tooluniverse-admet-predictiongit clone --depth 1 https://github.com/mims-harvard/ToolUniverseWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/mims-harvard/tooluniverse/tooluniverse-admet-prediction)<a href="https://agentmods.dev/skills/mims-harvard/tooluniverse/tooluniverse-admet-prediction"><img src="https://agentmods.dev/badge/skills/mims-harvard/tooluniverse/tooluniverse-admet-prediction/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/mims-harvard/tooluniverse/tooluniverse-admet-prediction"><img src="https://agentmods.dev/badge/skills/mims-harvard/tooluniverse/tooluniverse-admet-prediction.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00106 | $0.03991 |
| Opus 5 | $0.00053 | $0.01996 |
| Sonnet 5 | $0.00021 | $0.00798 |
| Haiku 4.5 | $0.00011 | $0.00399 |
Grade A, and why
tooluniverse-admet-prediction scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 310 lines — stays where its author put it; the contents beside it link to each section on GitHub.
ADMET Prediction & Drug Candidate Profiling
ADMET reasoning: a drug fails if it can't be absorbed, distributes to wrong tissues, isn't metabolized safely, or isn't excreted. Evaluate each property independently — good absorption doesn't compensate for liver toxicity. The ADME properties determine whether a compound reaches its target at therapeutic concentrations; toxicity determines whether it's safe to do so. Prioritize experimental data (T2) over computational predictions (T3) — ADMETAI predictions are screening tools, not definitive verdicts. When a FAIL is flagged in any toxicity category (hERG, AMES, DILI), treat it as program-limiting until wet-lab data refutes it.
LOOK UP DON'T GUESS: never assume SMILES, CID, or experimental LD50 values — always call PubChem to resolve compound identity before any ADMETAI or PubChemTox call.
Comprehensive pharmacokinetic and toxicity profiling integrating AI-based ADMET predictions, rule-based drug-likeness filters, and experimental benchmarks from curated databases.
When to Use This Skill
Triggers:
- "What are the ADMET properties of [compound]?"
- "Is [drug] likely to cross the blood-brain barrier?"
- "Predict the toxicity of this SMILES: ..."
- "Does [compound] violate Lipinski's rule of five?"
- "Assess the drug-likeness of [molecule]"
- "What are the CYP interactions for [drug]?"
- "Pharmacokinetic profile of [compound]"
- "Is [compound] orally bioavailable?"
- "What is the LD50 / hERG liability of [molecule]?"
Input: Drug name (e.g., "ibuprofen") OR SMILES string (e.g., "CC(C)Cc1ccc(cc1)C(C)C(=O)O")
Before You Run
ADMETAI tools run a local model, so they need the ml extra:
uv pip install 'tooluniverse[ml]'
Without it the tools still appear in tu list (the config loads) but fail at
call time with ADMETModel requires 'admet-ai' package. Run
tooluniverse-doctor to confirm which optional groups are installed.
Expected console noise — not errors. The first ADMETAI call loads PyTorch
and prints warnings such as missing-GPU / Trainer messages from
PyTorch Lightning, and TypedStorage is deprecated from PyTorch. These are
emitted by the underlying libraries during normal CPU inference. Predictions
are unaffected — do not report them to the user as failures and do not retry
the call because of them. Only treat output as a failure if the tool returns an
error field or no predictions.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 12d ago First seen · 310 lines · 106 tokens per session scan A 0bb7112aa5f8
tooluniverse-admet-prediction is a skill published in the GitHub repository mims-harvard/ToolUniverse (1,680 stars, last pushed 3d ago), licensed Apache-2.0. It adds 106 tokens to every session and 3,991 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
dfam-check
Measure mesh files against Design for Additive Manufacturing (DfAM) rules and report printability findings per process (FDM, SLS, SLA/DLP, metal PBF, MJF). Use when the user asks whether a part is printable, wants overhang/wall-thickness/support analysis of an .stl, .obj, .ply, or .3mf mesh, wants a build-orientation…
nanoresearch-writing
Draft a LaTeX research paper from all previous stage outputs.
construct-toy-examples
Generate and analyze simpler examples that satisfy both the assumptions and the conclusion of a theorem statement or subgoal. Use when you are stuck in reasoning and need simpler examples to regain traction, or when you want to see where the assumptions take effect and gain intuition.
obtain-immediate-conclusions
Derive immediate mathematical consequences from a theorem statement or subgoal. Use when starting a new problem, branch, or subgoal, or when cheap progress or a cleaner reformulation is needed before deeper proof search.
astro-dso-doc
Generates a complete, polished HTML documentation page, a processing checklist, an AstroBin post JSON, a PixInsight process icon set (XPSM), AND a ready-to-paste PixInsight project Description field for a deep-sky object (DSO) astrophotography project. Use this skill whenever the user mentions astrophotography, a DSO…
intermediate-outputs
Use this skill when working with circuit discovery in language models, mechanistic interpretability, activation patching, attribution patching, or Layer-wise Relevance Propagation (LRP) for neural network analysis.