| name | pubmed-literature-search |
| description | Search 36M+ citations from PubMed and PubMed Central. Advanced search with boolean operators, MeSH terms, author/journal filters, and full-text access.
|
PubMed Medical Literature Search
Search over 36 million biomedical citations from PubMed and PubMed Central using the NCBI E-utilities API. This skill enables AI agents to perform advanced literature searches with boolean operators, MeSH term expansion, author and journal filtering, date range restrictions, and direct full-text access links -- bringing the power of systematic literature review directly into clinical and research AI workflows.
Quick Install
npx skills add Open-Medica/open-medical-skills --skill pubmed-literature-search
What It Does
- Advanced boolean search across PubMed's entire 36M+ citation database using AND, OR, NOT operators with field-specific qualifiers (title, abstract, author, journal, MeSH)
- MeSH term integration for standardized medical vocabulary searches, including automatic term explosion to capture narrower descriptors in the MeSH hierarchy
- Citation metadata retrieval including authors, journal, publication date, DOI, PubMed ID, abstract text, MeSH headings, and publication type
- Full-text access links via PubMed Central (PMC) when articles are available open-access, with fallback to DOI resolver links
- Date-range and filter support to restrict results by publication year, article type (RCT, meta-analysis, review, case report), language, and species (human/animal)
Clinical Use Cases
- Systematic literature review: A researcher preparing a systematic review on SGLT2 inhibitors in heart failure can construct complex boolean queries with MeSH terms, retrieve all matching citations with abstracts, and export structured results for screening -- replacing hours of manual PubMed navigation.
- Point-of-care evidence lookup: A hospitalist encountering an unusual drug reaction can quickly search for case reports and pharmacovigilance studies linking the specific drug and adverse event, receiving summarized abstracts within seconds.
- Guideline development support: A clinical guideline committee can programmatically search for all RCTs and meta-analyses on a specific intervention, filter by date range and quality indicators, and generate a structured evidence table for committee review.
- Research gap identification: An academic physician planning a grant proposal can query existing literature on a topic, identify the most recent publications, and determine where evidence gaps remain to justify new research.
Safety & Evidence
- Safety Classification: Safe -- This skill queries a public, read-only database (NCBI PubMed). It does not make clinical decisions, alter patient data, or prescribe treatments. All returned data consists of published, peer-reviewed citations.
- Evidence Level: High -- PubMed is the gold-standard biomedical literature database maintained by the U.S. National Library of Medicine (NLM). The E-utilities API is the official programmatic interface used by thousands of research institutions worldwide.
Example Usage
Search for recent meta-analyses on metformin and cardiovascular outcomes:
Search PubMed for meta-analyses published in the last 5 years
examining metformin's effect on cardiovascular mortality in
type 2 diabetes. Use MeSH terms and restrict to English-language
articles.
Find case reports of a rare adverse drug reaction:
Search PubMed for case reports of pembrolizumab-induced
myocarditis. Include the abstract text and PMC full-text links
where available. Sort by most recent first.
Technical Details
- Category: Research
- Author: OMS Contributors (original: Augmented-Nature)
- License: MIT
- API: NCBI E-utilities (ESearch, EFetch, ESummary, ELink)
- Rate Limits: NCBI allows 3 requests/second without API key, 10 requests/second with a free NCBI API key
- Dependencies: HTTP client (curl/fetch), XML/JSON parser
References
This skill is part of Open Medical Skills, a curated marketplace of medical AI skills maintained by physicians for physicians and the healthcare industry.
