Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →brycewang-stanford/dataverse-api
Deposit and discover research datasets via Harvard Dataverse API
Passed 22 of 22 test cases with the skill, and 21 of 22 without it.
1 fixed21 passed either way0 still failing0 broke
Tokens, turns and time describe resource use, not a dollar price.
A test case the skill fixed. Same model, same prompt, with and without the skill.
I need an HTTP search query against Harvard Dataverse to retrieve genomics datasets published on or after January 1st, 2024. What field name and range syntax should be used in the `fq` parameter?
| Checked | Without skill | With skill |
|---|---|---|
| The date filter specifies the dateSort field. | Missed | Passed |
| The upper bound of the date range is unbounded using an asterisk. | Passed | Passed |
The instructions your model reads when the skill loads.
Dataverse is an open-source research data repository platform developed by Harvard IQSS, hosting 150K+ datasets across 80+ installations worldwide. The Harvard Dataverse alone has 130K+ datasets covering social science, natural science, and humanities. The API supports search, metadata retrieval, file download, and dataset deposit. Free, no authentication for read access.
https://dataverse.harvard.edu/api
# Search datasets
curl "https://dataverse.harvard.edu/api/search?q=climate+change&type=dataset&per_page=20"
# Search files within datasets
curl "https://dataverse.harvard.edu/api/search?q=temperature+data&type=file&per_page=20"
# Filter by subject
curl "https://dataverse.harvard.edu/api/search?q=survey+data&type=dataset&\
fq=subject_ss:\"Social Sciences\""
# Filter by publication date
curl "https://dataverse.harvard.edu/api/search?q=genomics&type=dataset&\
fq=dateSort:[2024-01-01T00:00:00Z TO *]"
# Sort by relevance or date
curl "https://dataverse.harvard.edu/api/search?q=machine+learning&type=dataset&\
sort=date&order=desc"
# By persistent ID (DOI)
test cases passed, vs 21 / 22 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.