ToolUniverse SDK — build and test agent tools
Build and test tools with the ToolUniverse framework — local tool registration, test-case generation, validation loops, remote sync and release via the ToolUniverse API — for anyone extending its open-source universe of 2 000+ tools across 40 domains
- What
- Build and test tools with the ToolUniverse framework — local tool registration, test-case generation, validation loops, remote sync and release via the ToolUniverse API — for anyone extending its open-source universe of 2 000+ tools across 40 domains
- Cost
- Free
- Needs
- a Python environment and the ToolUniverse repo checked out where the agent can add tool definitions — no key needed for local development; a ToolUniverse API key only for remote registration/release (save it in the secure vault, never paste a live key here)
- Install
- Copy the installer prompt below into your Muse — your agent does the rest.
Curated by Skill Harbor — @mims-harvard's ToolUniverse SDK skill: a TDD-style workflow for developing tools inside the open-source ToolUniverse framework (2 000+ agent tools across 40 domains). The agent's tasks: add/update local tool definitions with the `run`/`register_tool`/`register_remote_tool` APIs, generate test cases with `TestCaseGenerator` (from config docs or free-text requirements), validate with the interactive test loop (`iterate_and_validate`) — which spins up a ToolUniverse API server in `server_mode`, syncs the local tool, and iterates the LLM agent through failures until the tool passes — then release through the remote release flow with the ToolUniverse HTTP API (`/tools/register_remote_tool`, `/release`), requiring `TOOLUNIVERSE_API_KEY` for remote calls. Reference material covers the tool data model (definitions, arguments, results, error types), OpenAPI import, the HTTP API client, and the test-case generator config. Honest caveats: the core is open source (Apache-2.0) and local dev needs nothing but Python — an OpenAI key is only used by the LLM-search parts of the tool database, not by the SDK workflow itself; remote registration/release needs a ToolUniverse API key — save it in your secure vault, never paste a live key here. Apache-2.0 licensed. Skill Harbor never reviews the code, review it yourself before use. Discovered via skills.sh.
Version:
Install
Prerequisites: a Python environment and the ToolUniverse repo checked out where the agent can add tool definitions — no key needed for local development; a ToolUniverse API key only for remote registration/release (save it in the secure vault, never paste a live key here) Install "ToolUniverse SDK — build and test agent tools" for me. It gives my agent @mims-harvard's ToolUniverse development workflow: add or update local tool definitions (run/register_tool/register_remote_tool APIs), generate test cases with TestCaseGenerator (from config docs or plain-text requirements), validate through the iterate_and_validate loop (server_mode API server, local tool sync, iterate the LLM agent through failures until the tool passes), then release via the ToolUniverse HTTP API (/tools/register_remote_tool, /release) — with reference docs on the tool data model, OpenAPI import, the HTTP API client and test-case generator config. Apache-2.0-licensed. Repository: https://github.com/mims-harvard/tooluniverse/blob/main/plugin/skills/tooluniverse-sdk/SKILL.md 1. Fetch the SKILL.md file from the repository path into a temporary folder and summarize what it does in one or two sentences. 2. Safety check: review the SKILL.md for anything suspicious (unexpected network calls, shell commands, credential harvesting). This repo should contain zero secrets in code — the only credentials are TOOLUNIVERSE_API_KEY and optionally OPENAI_API_KEY as environment variables, never committed. Verify that holds here; STOP on any red flag and tell me. 3. Install it as a skill: copy SKILL.md and its helper files into the agent's skills directory, in a folder named "tooluniverse-sdk". 4. Verify with no network calls: frontmatter valid, files in place. 5. Report what was installed, where, and what I still need to do myself (e.g. clone the ToolUniverse repo and point the agent at my Python environment; save a ToolUniverse API key in the secure vault if I want remote registration/release — that's on me). GitHub is optional: if I have a GitHub account or the gh CLI, you may use it; otherwise public access is fine. Never require it unless it's in the prerequisites above. Rules: don't touch anything outside the temp folder and the install target. If anything looks off, stop and ask me.
Questions
How do I install a build?
Every product page includes a copy-paste install prompt. Paste it into your Muse and it sets the build up for you — no manual configuration.
Where does my money go?
Straight to the seller. Skill Harbor never processes payments: checkout happens on the seller’s own page, usually Stripe.
What does the ✓ next to a creator’s name mean?
It means we confirmed the identity of the person behind the listing. It says nothing about the code itself — always check a build before installing it.