tooluniverse-sdk

Build AI scientist systems with the ToolUniverse Python SDK for scientific research. Covers the 3 calling patterns (`tu.run` portable dict API, `tu.tools.X` function API, direct class instantiation), tool loading, batch execution, MCP server integration, and embedding-based tool search. Use for SDK programming, custom tool composition, benchmarking pipelines, and integrating ToolUniverse into research workflows.

Quality

86%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Quality

Content

80%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The content is concise and highly actionable with solid executable examples, but workflow clarity is capped by missing validation checkpoints in batch operations, and progressive disclosure is undercut by a referenced REFERENCE.md that does not exist in the bundle.

Suggestions

Add an explicit validation checkpoint after batch execution (e.g., check tu.run_batch results for per-call errors/status before proceeding) to lift workflow clarity above the batch-operation cap.

Either create the referenced REFERENCE.md with the detailed guides or remove/replace the [REFERENCE.md](REFERENCE.md) link so navigation does not dead-end.

Fold the load_tools() prerequisite into the Quick Start as a numbered, gated step ('1. load_tools() — REQUIRED before any tool call') so the critical workflow checkpoint is sequenced rather than only stated in notes.

Dimension	Reasoning	Score
Conciseness	The body is lean and code-first throughout — installation, quick start, batch, workflow, configuration, error handling, and a category table all earn their tokens, with no padding about concepts Claude already knows.	5 / 5
Actionability	Fully executable, copy-paste-ready code covers the common cases (tool finding, dict/function APIs, batch execution, caching/hooks, error handling), plus concrete accessor patterns like tu.all_tool_dict["ToolName"]['parameter'].get('required', []).	5 / 5
Workflow Clarity	The core sequence (init -> load_tools -> find -> execute -> batch) is present, and the load_tools() prerequisite plus error-handling recovery are noted, but batch operations lack explicit validate-after-step checkpoints, which the rubric caps at 3 for batch contexts.	3 / 5
Progressive Disclosure	Sections are well organized as an overview, but the single external reference ([REFERENCE.md](REFERENCE.md)) is a broken/dangling link — no references/, scripts/, or assets/ bundle and no REFERENCE.md exist — so the signaled navigation path dead-ends, which is more than a minor organization gap.	3 / 5
	Total	16 / 20 Passed

Description

92%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is strong: it enumerates concrete capabilities, provides an explicit 'Use for' trigger clause, and names a distinct, well-scoped SDK. The only soft spot is trigger-term coverage, which is good but not exhaustively synonym-rich.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete capabilities — 'the 3 calling patterns (tu.run portable dict API, tu.tools.X function API, direct class instantiation), tool loading, batch execution, MCP server integration, and embedding-based tool search' — giving comprehensive coverage of what the skill does.	5 / 5
Completeness	Explicitly answers 'what' (calling patterns, tool loading, batch, MCP, embedding search) and 'when' via a concrete 'Use for ...' clause listing SDK programming, tool composition, benchmarking pipelines, and research-workflow integration.	5 / 5
Trigger Term Quality	Includes natural trigger phrases ('SDK programming', 'custom tool composition', 'benchmarking pipelines', 'integrating ToolUniverse into research workflows') that users would plausibly say, but coverage lacks synonyms/variants a researcher might use, falling just below the comprehensive anchor.	4 / 5
Distinctiveness Conflict Risk	Names a specific product ('ToolUniverse Python SDK for scientific research') with distinct, narrow triggers, giving it a clear niche and minimal conflict risk with other skills.	5 / 5
	Total	19 / 20 Passed

Validation

93%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
relative_links	Relative link issues: 1 missing	Warning

	Total	15 / 16 Passed

Repository: mims-harvard/ToolUniverse
Path: plugin/skills/tooluniverse-sdk/SKILL.md
Commit: 1b02920

Reviewed: 1 day ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.