tooluniverse-dataset-discovery — скилл для Claude Code: как установить
Скилл tooluniverse-dataset-discovery находит научные датасеты под исследовательские задачи, проверяя дизайн исследования и переменные. Пригодится при поиске когорт, опросов или биомедицинских баз данных.
Apache-2.0 254 установок ★ 1 798 обновлён 2026-10-24
npx skills add mims-harvard/ToolUniverse --skill tooluniverse-dataset-discovery -a claude-codeЧто делает tooluniverse-dataset-discovery
Скилл подбирает научные репозитории под исследовательскую гипотезу. Сначала агент определяет минимальные требования к данным. Он выбирает дизайн исследования: лонгитюдный или одномоментный. Если гипотеза проверяет изменения во времени, скилл отсекает поперечные срезы. Они не фиксируют динамику. Также агент составляет список факторов риска и обязательных ковариат.
Поиск идёт по нескольким уровням: реестр re3data, базы вроде NHANES и GEO, а также разделы «Методы» в статьях PubMed. Скилл сразу сопоставляет размер выборки и тип доступа к файлам.
Для баз уровня dbGaP или EGA скилл выдаёт только метаданные исследований. Первичную генетическую информацию агент скачать не сможет без оформленной заявки на доступ. Для открытых источников агент сразу пишет скрипт на Python и забирает таблицы через Bash.
Какие задачи решает
- подобрать когорты под проверку причинно-следственных связей
- найти биомедицинские наборы данных в GEO, SRA или dbGaP
- оценить репрезентативность выборки и достаточность мощности исследования
- проверить наличие ключевых ковариат и конфаундеров в опросах NHANES
Примеры запросов агенту
- Найди лонгитюдные исследования связи дефицита сна и когнитивных нарушений.
- Где взять данные опросов NHANES по уровню сахара в крови с учётом возраста и пола?
- Проверь в GEO открытые наборы данных по профилям экспрессии генов при астме.
Как установить tooluniverse-dataset-discovery
1Через Skills CLI (Vercel)
npx skills add mims-harvard/ToolUniverse --skill tooluniverse-dataset-discovery -a claude-code2Вручную: положить файлы
~/.claude/skills/tooluniverse-dataset-discovery/SKILL.mdПапка скилла целиком; для одного проекта — .claude/skills/tooluniverse-dataset-discovery/
1Через Skills CLI (Vercel)
npx skills add mims-harvard/ToolUniverse --skill tooluniverse-dataset-discovery -a codex2Вручную: положить файлы
.agents/skills/tooluniverse-dataset-discovery/SKILL.mdДля всех проектов — ~/.agents/skills/
1Через Skills CLI (Vercel)
npx skills add mims-harvard/ToolUniverse --skill tooluniverse-dataset-discovery -a cursor2Вручную: положить файлы
.cursor/skills/tooluniverse-dataset-discovery/SKILL.mdCursor читает и .agents/skills/, и .claude/skills/
1GitHub CLI
gh skill install mims-harvard/ToolUniverse skills/tooluniverse-dataset-discovery2Вручную: положить файлы
.github/skills/tooluniverse-dataset-discovery/SKILL.md1Через Skills CLI (Vercel)
npx skills add mims-harvard/ToolUniverse --skill tooluniverse-dataset-discovery -a gemini-cli2Командой платформы
gemini skills install https://github.com/mims-harvard/ToolUniverse --path skills/tooluniverse-dataset-discoveryВручную: положить файлы
.devin/skills/tooluniverse-dataset-discovery/SKILL.mdСтарый путь .windsurf/skills/ тоже читается
Через Skills CLI (Vercel)
npx skills add mims-harvard/ToolUniverse --skill tooluniverse-dataset-discovery -a opencodeClaude.ai
Упакуйте папку в zip и загрузите в Settings → Capabilities → Skills
По шагам, с командами для каждой платформы: Как установить скилл в Claude Code: пять способов
Вопросы о tooluniverse-dataset-discovery
Как установить tooluniverse-dataset-discovery в Claude Code?
Выполните в терминале команду npx skills add mims-harvard/ToolUniverse --skill tooluniverse-dataset-discovery -a claude-code.
Какие репозитории охватывает этот скилл?
Он ищет по реестру re3data, архивам опросов CDC и NHANES, генетическим базам GEO, SRA, DDBJ, порталу Synapse, а также по публикациям в PubMed и ClinicalTrials.gov.
Позволяет ли скилл сразу скачать закрытые генетические данные?
Нет, для баз вроде dbGaP или EGA инструменты возвращают только метаданные исследований. Доступ к первичным генотипам требует отдельной официальной заявки исследователя.
Исходный текст
SKILL.md · Apache-2.0
---
name: tooluniverse-dataset-discovery
description: 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.
disable-model-invocation: true
---
# Dataset Discovery
## When to Use
- User asks "find me data about X" or "where can I get data on Y"
- User wants to analyze a relationship between variables
- User needs specific study designs (longitudinal, cross-sectional, experimental)
- User asks about specific surveys or cohorts
## Step 1: Understand What the Research Question Requires
Before searching, determine the **minimum data requirements**:
**Study design needed:**
- "Does X predict CHANGES in Y over time?" → longitudinal (same people measured repeatedly). Cross-sectional data CANNOT answer this — don't settle for it.
- "Is X associated with Y?" → cross-sectional is sufficient (one-time measurement)
- "Does intervention X cause outcome Y?" → experimental (clinical trial with controls)
- "What genes/proteins are involved in X?" → omics (sequencing, expression, proteomics)
**Variables needed:**
- List the specific exposure, outcome, and confounder variables
- For each variable, note the measurement type (continuous, categorical, biomarker vs self-report)
- Identify minimum confounders needed (age, sex are almost always required; domain-specific confounders depend on the question)
**Population needed:**
- Age range, geography, clinical status, sample size requirements
- Power analysis: to detect a small effect (r=0.1), you need ~800 subjects at 80% power
## Step 2: Search Strategy
Search from broadest to most specific. Use `find_tools` to discover available dataset search tools — don't rely on memorized tool names.
**Layer 1 — Cross-repository search (cast wide net):**
Search tools that index datasets across thousands of repositories. These find datasets you didn't know existed.
- Search by: research topic keywords, variable names, population descriptors
- Look for: DOI-registered datasets, repository listings, government data portals
- `re3data_search_repositories`/`re3data_get_repository` — a registry of 3,000+ research data REPOSITORIES themselves (not individual datasets), covering every discipline; use this one level up from a dataset search when the question is "what repository should I even be looking in for this kind of data" (verified live: searching returns real health-research repositories, e.g. the "EPIC study" cohort and "Pacific Islands Families Study"). Once you've identified the right repository from re3data, go search *inside* it with the domain-specific tools in Layer 2 below.
**Layer 2 — Domain-specific repositories:**
Search repositories specialized for your data type.
- Health surveys: CDC, NHANES (US; search by variable name, not topic keywords); DHS Program (`DHSProgram_search_indicators`/`_get_data`) for the international/LMIC equivalent -- fertility, maternal/child mortality, nutrition, immunization, HIV indicators across national surveys, filterable by country and year
- Genomics (open): SRA, ENA, ArrayExpress, GEO, and DDBJ (`DDBJ_search_entries`/`_get_entry`/`_get_cross_references`) -- the third INSDC archive alongside SRA/ENA, plus GEA/MetaboBank entries; also GSA (`GSA_get_accession`) for China's national genomics archive
- Genomics (controlled-access, human studies): dbGaP (`DbGaP_search_studies`/`_get_study`) and EGA (`EGA_get_study`/`_get_dataset`/`_get_study_datasets`) -- both require a separate data-access request process; these tools return study/dataset metadata, not the underlying genotype data itself
- Proteomics: PRIDE, MassIVE
- Metabolomics: MetaboLights, Metabolomics Workbench
- Clinical: ClinicalTrials.gov (for trial data with results)
- Consortium research portals (AD Knowledge Portal, NF-OSI, PsychENCODE, HTAN, DREAM challenges and others): `Synapse_search_entities` finds public Synapse projects/folders/files/tables/datasets by keyword (`node_type` narrows it), `Synapse_get_entity` returns the curated annotations (assay, species, tissue, disease), and `Synapse_list_children` browses a project. Anonymous and read-only: restricted entities return a clear "not public" error and file contents are not downloaded -- data access still goes through synapse.org.
**Layer 3 — Literature-based discovery:**
Many datasets aren't in any repository — they're described in paper methods sections.
- Search PubMed/EuropePMC for papers that analyzed the relationship you're interested in
- Read their methods: "We used data from [DATASET NAME]" tells you exactly what exists
- Check supplementary materials for deposited data (GEO/SRA accession numbers)
- This is often the MOST effective strategy for finding niche datasets
## Step 3: Evaluate Dataset Fitness
For each candidate dataset, assess these dimensions:
**Variables:**
- Does it contain your SPECIFIC exposure and outcome variables?
- Are they measured the way you need? (biomarker vs self-report, continuous vs categorical)
- Are key confounders available? (missing confounders = biased analysis)
**Design match:**
- If you need longitudinal: does it follow the SAME individuals over time? How many waves? What's the follow-up interval?
- Beware: "repeated cross-sections" (different people each wave) are NOT longitudinal
- If you need experimental: is there a proper control group? Randomization?
**Sample:**
- Is the sample large enough for your analysis? (logistic regression needs ~10 events per predictor)
- Does the population match? (age range, geography, clinical characteristics)
- Are there subgroups you need? (stratified by sex, race, disease status)
**Access:**
- Publicly downloadable (best) vs registration required (days) vs collaboration agreement (months) vs restricted (may be impossible)
- Data format: CSV/TSV (easy), XPT/SAS (need conversion), proprietary database (may need special software)
**Quality:**
- Is it from a well-known study with published methods? (NHANES, HRS, UK Biobank = high quality)
- Has it been used in peer-reviewed publications? (indicates data is usable)
- What's the response rate / missingness pattern?
## Step 4: Download and Analyze
Don't stop at finding datasets — download and analyze them. Write and run Python code via Bash. Never describe what you "would do" — execute it.
### Data Loading Cookbook
Choose the loader that matches your data source. When unsure of the format, download a small sample first and inspect.
```python
import requests, io, pandas as pd
# --- Tabular files (most common) ---
df = pd.read_csv("data.csv") # CSV / TSV (use sep="\t" for TSV)
df = pd.read_excel("data.xlsx") # Excel
df = pd.read_stata("data.dta") # Stata
df = pd.read_sas("data.xpt", format="xport") # SAS transport (XPT)
df = pd.read_sas("data.sas7bdat", format="sas7bdat") # SAS native
df = pd.read_parquet("data.parquet") # Parquet
df = pd.read_json("data.json") # JSON (records or columnar)
df = pd.read_fwf("data.dat") # Fixed-width (some legacy surveys)
# --- Download from URL first, then parse ---
resp = requests.get(url, timeout=120)
content = resp.content
# Detect format from URL or content header
if url.endswith(".XPT") or url.endswith(".xpt"):
df = pd.read_sas(io.BytesIO(content), format="xport")
elif url.endswith(".csv") or url.endswith(".csv.gz"):
df = pd.read_csv(io.BytesIO(content))
elif url.endswith(".tsv") or url.endswith(".tsv.gz"):
df = pd.read_csv(io.BytesIO(content), sep="\t")
elif url.endswith(".json"):
df = pd.read_json(io.BytesIO(content))
else:
# Try CSV first, then inspect
df = pd.read_csv(io.BytesIO(content))
# --- REST API pagination (common for GDC, ClinicalTrials.gov, etc.) ---
import json
all_records = []
offset = 0
while True:
resp = requests.get(f"{api_url}?offset={offset}&limit=100", timeout=30)
batch = resp.json().get("data", [])
if not batch:
break
all_records.extend(batch)
offset += len(batch)
df = pd.DataFrame(all_records)
```
### Merge, Clean, Analyze
```python
# Merge multiple files on participant/sample ID
merged = df1.merge(df2, on="id_col", how="inner")
# Filter population
subset = merged[(merged["age"] >= 60) & (merged["age"] <= 80)].copy()
# Handle missing values
missing_pct = subset.isnull().mean() * 100
print("Missing % per variable:\n", missing_pct[missing_pct > 0].sort_values(ascending=False))
subset = subset.dropna(subset=["exposure_var", "outcome_var"])
# Quick regression
import statsmodels.formula.api as smf
model = smf.ols("outcome ~ exposure + age + sex", data=subset).fit()
print(model.summary())
# Visualization
import matplotlib.pyplot as plt
plt.scatter(subset["exposure"], subset["outcome"], alpha=0.3)
plt.xlabel("Exposure"); plt.ylabel("Outcome")
plt.savefig("/tmp/scatter.png", dpi=150, bbox_inches="tight")
```
Always run the code and report actual numbers (β, p-value, CI, N).
## Step 5: Report Honestly
Structure the report as:
1. **Best available dataset** — name, what it contains, access method, key limitation
2. **Analysis results** — actual statistics (β, p-value, CI, N) from running the code
3. **Alternative datasets** — ranked by fitness, with tradeoffs
4. **What CANNOT be answered** — if no dataset matches the study design needed, say so clearly
5. **Recommended next steps** — apply for access to longitudinal data, replicate in other cohorts
**Critical honesty rules:**
- Never claim a dataset answers a temporal question if it's cross-sectional
- Distinguish "data exists but needs registration" from "data doesn't exist"
- Report actual computed statistics, not hypothetical analyses
- State the strongest analysis possible with available data, even if it's weaker than what was asked
## LOOK UP, DON'T GUESS
Never assume a dataset exists — search for it. Never assume access is public — check. Never assume variables are measured the way you need — verify the codebook.
© mims-harvard · Apache-2.0 · mims-harvard/ToolUniverse