What does this skill do?

The Biomedical Search on PubMed Skill provides a structured workflow for searching biomedical literature using the MeSH controlled vocabulary and the NCBI E-utilities API. It allows you to build advanced queries with field tags and filters by publication type, date, and language, and to retrieve PMIDs, abstracts, and metadata in an automated and reproducible manner.

Systematic Reviews
Perform search runs using reusable strings that include MeSH terms, Boolean operators, and article type filters.
Advanced MeSH Searches
Combine MeSH terms with free-text from the title or abstract, and narrow your search by date, language, or publication type.
Retrieval via API
Use the E-utilities endpoints `esearch`, `esummary`, and `efetch` to retrieve PMIDs, light metadata, and full abstracts in XML/MEDLINE format.
Integration into Scripts
Integrate PubMed searches into Python scripts or other HTTP clients by configuring your email and API key in environment variables.

Usage examples

🔬 Filtered MeSH Search
Search for clinical trials on metformin for type 2 diabetes published between 2020 and 2026 in English, and retrieve the abstracts.
📋 Systematic Review
Build a MeSH query for systematic reviews on hypertension and run it via E-utilities, returning the PMIDs in JSON.
🧬 Thematic Search
Retrieve the PMIDs for articles on CRISPR and sickle cell disease, and generate a search log that includes the date, query, and number of results.
🐍 Python script
Create a Python script that searches for literature on breast cancer using E-utilities and returns the metadata in JSON format.

Features

Controlled MeSH Vocabulary Prioritize MeSH terms for established concepts and combine them with free-text terms when the topic is emerging.
Advanced Field Labels It supports tags such as [ti], [ab], [mh], [pt], [dp], and [la] to narrow down searches with precision.
Integration with NCBI E-utilities Use the esearch, esummary, and efetch endpoints to automate the retrieval of PMIDs, metadata, and abstracts.
Reproducible Search Log Generate a record containing the exact string, database, date, applied filters, and number of results returned.
Settings by Environment This allows you to define the NCBI email and API key in environment variables to comply with the API rate limits.

Frequently asked questions

It is not required, but it is highly recommended. Without an API key, you are limited to 3 requests per second; with it, you can increase that to 10. Set it in the NCBI_API_KEY environment variable.
Yes. The skill is designed to generate repeatable search strings using MeSH terms, Boolean operators, and filters, and it produces a search log containing all the parameters to ensure reproducibility.
esearch returns PMIDs in JSON; esummary provides lightweight metadata; and efetch allows you to retrieve full abstracts and records in XML/MEDLINE format.
Yes. Claude builds MeSH queries and can run them for you. The Python script is an example of integration, but the skill guides the query construction regardless of the language.
Biomedical Search on PubMed — Scientific Literature with Claude AI

¿Prefieres escuchar el contenido? Genera la narración de audio con un clic.