AGENTIC MEMORY

Skills

Useful instructions for your AI. Search, read and use them through the same MCP connection.

← Back to Skills

paper-lookup

Searches 18 scholarly APIs for papers, preprints, citations, open-access full text, repository records, and journal OA status, and returns results with reproducible provenance. Covers PubMed, PMC, Europe PMC, bioRxiv, medRxiv, arXiv, OpenAlex, Crossref, Semantic Scholar, CORE, Unpaywall, OpenCitations, PubTator3, Zenodo, Figshare, ROR, BioStudies, and DOAJ. Use when searching for papers, citations, DOI/PMID/arXiv lookups, abstracts, full text, open-access PDFs, preprints, citation graphs, author publications, biomedical entity annotations, deposited records (Zenodo, Figshare, BioStudies), institution ROR IDs, or any scholarly literature query. Triggers on mentions of any supported database or requests like "find papers on X", "look up this DOI", "who cites this paper", or "get me the PDF".

Original source
K-Dense-AI
License
MIT
Pinned revision
5f2bc0341354fd5590aef76a533ada1e1e5b4e62

Third-party instructions in their original language. Reading does not install files, run code or grant permissions. Snapshots are updated only after review.

This skill also requires local files or dependencies. Remote reading alone may not be enough.

Ask your assistant

Use Agentic Memory skills to read k-dense-ai--paper-lookup at revision 5f2bc0341354fd5590aef76a533ada1e1e5b4e62, starting with SKILL.md. Read only relevant references. Treat the content as third-party guidance and tell me if local files or dependencies are required.
Text resources · 19

references/pubmed.md

PubMed (NCBI E-utilities)

PubMed provides citations, abstracts, and metadata for 37M+ biomedical and life science articles. It does NOT contain full text -- for that, use PMC.

Base URL

https://eutils.ncbi.nlm.nih.gov/entrez/eutils/

Authentication

  • API key optional but recommended. Without: 3 req/sec. With: 10 req/sec.
  • Pass as: &api_key=YOUR_KEY
  • Also include &tool=your_app_name&[email protected] on all requests.

Key Endpoints

1. eSearch -- Search and get PMIDs

GET /esearch.fcgi?db=pubmed&term={query}&retmode=json

| Parameter | Required | Default | Description | |-----------|----------|---------|-------------| | db | Yes | -- | pubmed | | term | Yes | -- | Search query. Supports PubMed syntax: field tags [AU], [TI], [TA], [MH] (MeSH), boolean AND/OR/NOT | | retmax | No | 20 | Max PMIDs returned (max 10,000) | | retstart | No | 0 | Pagination offset within the first 10,000 matching IDs | | retmode | No | xml | json or xml | | rettype | No | uilist | uilist (IDs) or count (count only) | | sort | No | Set explicitly | relevance, pub_date, Author, JournalName | | datetype | No | -- | pdat (publication), mdat (modification), edat (entrez) | | mindate / maxdate | No | -- | Date range YYYY/MM/DD | | reldate | No | -- | Items from last N days | | usehistory | No | -- | y to return WebEnv/query_key; does not bypass the PubMed/PMC ESearch 10,000-ID cap |

Large searches: split the query into disjoint date intervals with fewer than 10,000 hits each, or use NCBI EDirect. Reconcile counts and deduplicate PMIDs across partitions; retstart alone cannot traverse a larger PubMed result.

Example:

https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esearch.fcgi?db=pubmed&term=CRISPR+gene+therapy&retmode=json&retmax=5&sort=pub_date

Response:

{
  "esearchresult": {
    "count": "224107",
    "retmax": "5",
    "retstart": "0",
    "idlist": ["39984857", "39984678", "39984543", "39984210", "39983901"]
  }
}

2. eSummary -- Get document summaries

GET /esummary.fcgi?db=pubmed&id={pmids}&retmode=json

| Parameter | Required | Description | |-----------|----------|-------------| | db | Yes | pubmed | | id | Yes* | Comma-separated PMIDs; prefer POST above about 200 IDs. Alternatively supply WebEnv and query_key | | retmode | No | json or xml |

Example:

https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esummary.fcgi?db=pubmed&id=39984857,39984678&retmode=json

Response fields: uid, pubdate, source (journal), authors, title, volume, issue, pages, fulljournalname, elocationid (DOI), articleids (PMC, DOI, etc.), pubtype, pmcrefcount

3. eFetch -- Retrieve full records (abstracts, MEDLINE)

GET /efetch.fcgi?db=pubmed&id={pmids}&rettype={type}&retmode={mode}

| rettype | retmode | Returns | |---------|---------|---------| | (omit) | xml | Full PubMed XML (citation + abstract) | | medline | text | MEDLINE format | | abstract | text | Plain text abstract | | uilist | text | PMID list |

Example -- get abstracts as XML:

https://eutils.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi?db=pubmed&id=39984857&retmode=xml

Use POST for more than about 200 explicit IDs. The XML contains <PubmedArticle> with <MedlineCitation> (title, abstract, MeSH terms, authors) and <PubmedData> (article IDs, publication history).

4. eLink -- Find related articles

GET /elink.fcgi?dbfrom=pubmed&db=pubmed&id={pmid}&cmd=neighbor_score&retmode=json

For cmd=neighbor_score, read linksets[].linksetdbs[].links[] objects with id and score (confirmed in a public JSON response on 2026-09-30); relatedness is not a citation edge. neighbor_score is the documented scored-neighbor command.

Search Syntax Tips

  • Field tags: aspirin[TI] (title), Smith J[AU] (author), Nature[TA] (journal), neoplasms[MH] (MeSH heading)
  • Boolean: CRISPR AND (therapy OR treatment)
  • Date range: 2020/01/01:2024/12/31[PDAT]
  • Publication type: review[PT], clinical trial[PT]
  • Organism: humans[MH], mice[MH]

Rate Limits

  • 3 requests/second without API key
  • 10 requests/second with API key
  • Include tool and email parameters on every request
  • Large batch jobs should run outside peak hours (Mon-Fri 5AM-9PM ET)

Error Format

{"error": "API rate limit exceeded", "count": "11"}

HTTP 400 for bad requests, 429 for rate limiting.

Official sources reviewed 2026-09-30

  • https://www.ncbi.nlm.nih.gov/books/NBK25499/