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 create-tooluniverse-skillgit 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/create-tooluniverse-skill)<a href="https://agentmods.dev/skills/mims-harvard/tooluniverse/create-tooluniverse-skill"><img src="https://agentmods.dev/badge/skills/mims-harvard/tooluniverse/create-tooluniverse-skill/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/create-tooluniverse-skill"><img src="https://agentmods.dev/badge/skills/mims-harvard/tooluniverse/create-tooluniverse-skill.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- Socket pass
- Snyk pass
- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high Rogue Agent · line 35 Skill modifies its own code, configuration, or behavior at runtime. Self-modification enables an agent to escalate privileges, disable safety constraints, or install persistent backdoors.Fix: Prevent the skill from modifying its own code, SKILL.md, or configuration files. Treat skill files as read-only at runtime.
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 | $0.00023 | $0.01301 |
| Opus 5.5 | $0.00009 | $0.00520 |
| Sonnet 5.5 | $0.00005 | $0.00260 |
| Haiku 4.5 | $0.00002 | $0.00130 |
Grade A, and why
create-tooluniverse-skill 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 5d 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 — 120 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Create ToolUniverse Skill
Systematic workflow for creating production-ready ToolUniverse skills.
The
devtu-*skills this one hands off to are no longer in this repository.devtu-optimize-skills,devtu-create-toolanddevtu-fix-toolmoved to the tu-harness plugin (tooluniverse/tooluniverse-harness), because they operate on this repository rather than being something a user of it calls. Install them withclaude plugin install tu-harness@tooluniverse-harness; the references below name them by skill, so they resolve once that plugin is present. This skill is maintainer-facing too -- it is not in the user-facing plugin mirror -- and belongs beside them; it has not moved yet only because nothing forced the question.
Core Principles
Build on the 10 pillars from devtu-optimize-skills:
- TEST FIRST - never document untested tools
- Verify tool contracts - don't trust function names
- Handle SOAP tools - add
operationparameter - Implementation-agnostic docs - no Python/MCP code in SKILL.md
- Foundation first - query aggregators before specialized tools
- Disambiguate carefully - resolve IDs properly
- Implement fallbacks - Primary -> Fallback -> Default
- Grade evidence - T1-T4 tiers on claims
- Quantified completeness - numeric minimums per section
- Synthesize - models and hypotheses, not just lists
See OPTIMIZE_INTEGRATION.md for detailed application of each pillar.
7-Phase Workflow
| Phase | Duration | Description |
|---|---|---|
| 1. Domain Analysis | 15 min | Understand use cases, data types, analysis phases |
| 2. Tool Discovery | 30-45 min | Search, read configs, test tools (MANDATORY) |
| 3. Tool Creation | 0-60 min | Create missing tools via devtu-create-tool |
| 4. Implementation | 30-45 min | Write python_implementation.py with tested tools |
| 5. Documentation | 30-45 min | Write SKILL.md (agnostic) + QUICK_START.md |
| 6. Validation | 15-30 min | Run test suite, validate checklist, manual verify |
| 7. Packaging | 15 min | Create summary, update tracking |
What ships with it
18 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- assets/skill_template/python_implementation.py 5.5 KB runs code
- assets/skill_template/QUICK_START.md 7.5 KB
- assets/skill_template/SKILL.md 6.2 KB
- assets/skill_template/test_skill.py 6.6 KB runs code
- CODE_TEMPLATES.md 4.5 KB
- OPTIMIZE_INTEGRATION.md 3.3 KB
- PACKAGING_TEMPLATE.md 1.1 KB
- PARAMETER_VERIFICATION.md 1.1 KB
- QUICKSTART_TEMPLATE.md 1.6 KB
- references/devtu_optimize_integration.md 8.5 KB
- references/implementation_agnostic_format.md 6.3 KB
- references/skill_standards_checklist.md 7.1 KB
- references/tool_testing_workflow.md 5.2 KB
- scripts/test_tools_template.py 5.7 KB runs code
- SKILL_TEMPLATE.md 2.3 KB
- test_tools_template.py 5.7 KB runs code
- TESTING_GUIDE.md 6.3 KB
- VALIDATION_CHECKLIST.md 1.4 KB
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.
- 5d ago Changed · +12 lines 745998d4d3e6
- 1mo ago First seen · 108 lines · 23 tokens per session scan A 17e241f50b68
create-tooluniverse-skill is a skill published in the GitHub repository mims-harvard/ToolUniverse (1,757 stars, last pushed yesterday), licensed Apache-2.0. It adds 23 tokens to every session and 1,301 once invoked, about $0.0001 per session on Opus 5.5. 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 2026-09-03.
Other skills, from other repositories
azure-cosmos-db-py
Build Azure Cosmos DB NoSQL services with Python/FastAPI following production-grade patterns. Use when implementing database client setup with dual auth (DefaultAzureCredential + emulator), service layer classes with CRUD operations, partition key strategies, parameterized queries, or TDD patterns for Cosmos. Triggers…
resolve-author-fix
Find the root cause, implement a focused fix, and prove behavior with targeted tests and recordings.
testing
TDD/BDD testing principles. Use when writing tests, reviewing test coverage, setting up testing, or discussing test strategy and test architecture.
agent-test-generator
Specialist subagent: Generates comprehensive test suites using TDD patterns. Use only when the user explicitly asks for tests, improved coverage, or test-first development.
test-driven-development
Writing code with tests first: RED-GREEN-REFACTOR, reproduce bugs as failing tests before fixing, and avoid testing anti-patterns. Use when implementing behaviour, fixing a defect, or asked to add tests.
tdd
Test-driven development, from unit to end-to-end tests. Use when the user wants to build features or fix bugs test-first, mentions "red-green-refactor", or wants integration or end-to-end tests (Playwright, Cypress). Don't use to backfill tests on existing code (/refactor) or diagnose an unknown bug (/debug).