CtrlK
BlogDocsLog inGet started
Tessl Logo

tooluniverse-dataset-discovery

Find and evaluate research datasets for any scientific question. Maps research questions to required study designs (longitudinal vs cross-sectional, observational vs experimental, single-cohort vs multi-cohort). Use when the user asks 'find data about X', 'where can I get data on Y', or needs a specific cohort/survey/repository. Covers GEO, ArrayExpress, dbGaP, NHANES, UK Biobank, ClinicalTrials.gov, GWAS Catalog, and 30+ scientific repositories.

72

Quality

89%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Low

Low-risk findings worth noting

The canonical home for this skill is tooluniverse-dataset-discovery in mims-harvard/ToolUniverse

SKILL.md
Quality
Evals
Security

Quality

Content

78%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.

A strong, action-oriented skill with executable code and a clear five-step workflow. Its main weaknesses are a verbose inline data-loading cookbook that could be offloaded to a reference file and the absence of explicit validate-and-retry feedback loops for batch operations.

Suggestions

Move the Data Loading Cookbook (the multi-format pandas loader reference) into a references/DATA_LOADERS.md file and link to it from Step 4 to reduce inline tokens.

Add an explicit validate-then-fix feedback loop in Step 4 (e.g. 'run the loader on a small sample; if parsing fails, inspect the raw bytes/header, adjust the loader, retry').

Tighten prose that restates general knowledge (e.g. 'import json', what CSV/TSV are) so every token earns its place.

DimensionReasoningScore

Conciseness

Mostly efficient with actionable bullets and copy-paste code, but the large inline data-loading cookbook (multiple file formats) and merged example over-explain details Claude already knows about pandas loaders.

4 / 5

Actionability

Fully executable, copy-paste-ready Python covering download, format detection, pagination, merge/clean, regression, and plotting, with concrete decision rules (e.g. 'logistic regression needs ~10 events per predictor').

5 / 5

Workflow Clarity

Clear five-step sequence with ordered phases and validation-style checkpoints (e.g. 'verify the codebook', 'Never describe what you would do — execute it'), but lacks an explicit validate-then-fix feedback loop for batch data operations.

4 / 5

Progressive Disclosure

Well-organized with clear section headers, but all content is inlined into a single SKILL.md (no bundle files exist) and the cookbook/API-style material would benefit from being split into referenced files.

3 / 5

Total

16

/

20

Passed

Description

100%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 exemplary: concrete actions, explicit trigger phrases, a named-coverage list, and third-person voice. It clearly answers both what the skill does and when to invoke it with minimal conflict risk.

DimensionReasoningScore

Specificity

Lists multiple concrete actions ('Find and evaluate research datasets', 'Maps research questions to required study designs') plus a comprehensive coverage list (GEO, ArrayExpress, dbGaP, NHANES, UK Biobank, ClinicalTrials.gov, GWAS Catalog, 30+ repositories).

5 / 5

Completeness

Explicitly answers what ('Find and evaluate research datasets...') and when ('Use when the user asks...') with concrete trigger phrases — both halves present and specific.

5 / 5

Trigger Term Quality

Includes natural user phrasings ('find data about X', 'where can I get data on Y') alongside specific cohort/survey/repository triggers, matching how users actually ask for datasets.

5 / 5

Distinctiveness Conflict Risk

Clear niche (scientific dataset discovery with named repositories and study-design mapping) with distinct triggers; minimal overlap risk with adjacent skills.

5 / 5

Total

20

/

20

Passed

Validation

100%

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

Validation16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository
mims-harvard/ToolUniverse
Reviewed

Table of Contents

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.