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-connectgit 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-connect)<a href="https://agentmods.dev/skills/mims-harvard/tooluniverse/tooluniverse-connect"><img src="https://agentmods.dev/badge/skills/mims-harvard/tooluniverse/tooluniverse-connect/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-connect"><img src="https://agentmods.dev/badge/skills/mims-harvard/tooluniverse/tooluniverse-connect.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00138 | $0.02030 |
| Opus 5.5 | $0.00055 | $0.00812 |
| Sonnet 5.5 | $0.00028 | $0.00406 |
| Haiku 4.5 | $0.00014 | $0.00203 |
Grade C, and why
tooluniverse-connect scanned grade C with 2 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 yesterday.
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.
Downloads and executes remote codehighSupply chain
curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.
`claude mcp add --transport http --scope user tooluniverse "https://mcp.aiscientist.tools" && curl -fsSL https://connect.aiscientist.tools/skills.sh | sh -s -- claude-code && claude mcp login tooluniverse` Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
`claude mcp add --transport http --scope user tooluniverse "https://mcp.aiscientist.tools" && curl -fsSL https://connect.aiscientist.tools/skills.sh | sh -s -- claude-code && claude mcp login tooluniverse` How it starts
The opening of the file, as written. The whole thing — 85 lines — stays where its author put it; the contents beside it link to each section on GitHub.
ToolUniverse Connect: what each part is for and how to use it
The person you are helping is usually a scientist, not a programmer. Answer in plain words, one step at a time. When you can do a step yourself (run a command, call a tool), do it; ask the user only for what only they can do, such as signing in or clicking Allow.
Never ask the user to paste a key, token or password into the chat, and never write one into a file or command yourself.
1. Find out which ToolUniverse this user has
| What you see | The user is on |
|---|---|
ToolUniverse tools such as what_can_i_use, list_skills or get_skill, from the server https://mcp.aiscientist.tools (earlier setups: https://tooluniverse-backend.onrender.com/mcp, which still works) |
ToolUniverse Connect (hosted) |
A ToolUniverse server started on this computer (uvx tooluniverse, tu serve), the tu command, or Python from tooluniverse import ToolUniverse |
Local ToolUniverse |
| No ToolUniverse tools at all | Not connected yet |
To check from a terminal: claude mcp list (Claude Code) or codex mcp list (Codex). In Claude (claude.ai or the desktop app), a connector must also be switched on in each chat; see section 3.
2. Hosted or local: which one to suggest
| ToolUniverse Connect (hosted) | Local ToolUniverse | |
|---|---|---|
| Setup | Nothing to install: add one address to the AI app and sign in once in the browser | Install uv, run uvx tooluniverse, edit the AI app's settings |
| Database logins (API keys) | Saved once on the website, encrypted, and used in every connected app | Set as environment variables, separately in each app's settings |
| Tools | Most ToolUniverse tools. Tools that read or write files on a computer or run code are not offered | All tools, including file and code tools |
| Share tools with colleagues, use a colleague's GPU computer, publish a tool | Yes | No |
| Where requests go | Through the ToolUniverse Connect servers, then to each database | Straight from this computer to each database |
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.
- yesterday First seen · 85 lines · 138 tokens per session scan C 23824cafa93a
tooluniverse-connect is a skill published in the GitHub repository mims-harvard/ToolUniverse (1,744 stars, last pushed yesterday), licensed Apache-2.0. It adds 138 tokens to every session and 2,030 once invoked, about $0.0006 per session on Opus 5.5. A static security scan graded it C with 2 findings (downloads and executes remote code, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-16.
Other skills, from other repositories
ai-agent-tool-routing
Use when designing or reviewing an AI agent tool registry or MCP router: document parameter schemas, caller authorization, execution limits, replay-safe retries and verification evidence.
agent-engineering-expert
Build LLM agents that use tools safely: tool design, the agent loop, memory, MCP servers, multi-agent orchestration, sandboxing and prompt-injection defence. Use when the user mentions AI agents, tool use or function calling, MCP or Model Context Protocol, autonomous workflows, multi-agent systems, LangChain or…
external-repo-setup
Protocol for GPU agents to clone external GitHub repos, install their dependencies, download pretrained weights, and integrate them as featurizers or models inside a focus-area train.py. Use this when a proposal references a GitHub repo or pretrained checkpoint that is not already present in the focus-area workspace.
multi-agent-focus
Generic skill for self-organizing multi-agent teams that collaborate on an optimization problem. Agents discuss dimensions, form teams, run experiments, and adapt when stagnating. Uses AnonAPI posts for discussion and workspaces for shared state.
agentic-patterns
Fundamental patterns for effective agentic behavior. Teaches decomposition, tool orchestration, error recovery, context management, quality self-assessment, and knowing when to stop. Model-agnostic principles that make any agent more effective regardless of domain. Activate on: "how should I structure this agent"…
ai-agent-tool-routing
Autonomous AI agent tool router register: schema registration, runtime parameter coercion, idempotency keys, and error recovery policies.