clinical-trials-database

The clinical-trials-database skill provides read-only access to worldwide clinical trial data from ClinicalTrials.gov via its REST API v2. It solves the problem of querying a large public research database without hand-constructing raw API expressions, by wrapping the API in a CLI script (scripts/clinical_trials_api.py) that exposes dedicated flags for common filters such as condition, intervention, status, phase, age group, study type, and sponsor. Typical uses include searching trials by disease or drug, retrieving a trial by NCT ID, checking eligibility criteria, counting trials, mapping a sponsor's portfolio, and finding recruiting trials for patient matching.

The skill emphasizes efficient, well-behaved API usage: it instructs always using the provided wrapper (which enforces the required rate limit), always restricting responses with the --fields parameter because trial JSON records are large, checking --count-total before fetching all records, and paginating large result sets with --limit and --page-token. Reference documentation details the endpoints, query versus filter parameters, valid enum values, Essie expression syntax for advanced filtering, and the protocolSection/resultsSection response structure, with a schema reference for exact field paths. As a prerequisite it uses the uv tool to run the script, and it requires notifying the user to check the ClinicalTrials.gov terms and recording that notification in a local license file before use.

It targets researchers, medical and pharmaceutical analysts, and clinicians doing patient-trial matching or portfolio analysis. All behavior is read-only querying of a public government API with rate limiting and terms-of-use acknowledgement.