| name | pubmed-search |
| description | Search PubMed biomedical literature for a topic, gene, drug, disease, or clinical question and retrieve article details. Use when the user wants peer-reviewed biomedical or clinical references. Uses the plugin's built-in read-only PubMed MCP tool for discovery — no credential or install needed. |
| license | Apache-2.0 |
PubMed Literature Search (Codex-native)
Codex-native biomedical literature search wired to this plugin's built-in
read-only MCP tool. Inspired by Google DeepMind's pubmed-database
(science-skills, Apache-2.0);
the MCP integration is Codex Science's own.
How to run
- Search with the plugin MCP tool
science_search_pubmed:
- arguments:
query (topic/gene/drug/disease terms, ≤500 chars), limit
(1–10, default 5).
- returns a JSON list of
{id, title, url} where id is the PMID and
url is the PubMed entry. Note: this discovery tool returns PMIDs and
links only — no abstract.
- Fetch details for the PMIDs you keep: open each
url with a fetch tool,
or use NCBI E-utilities efetch/esummary
(https://eutils.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi?db=pubmed&id=<PMID>&retmode=xml)
to get title, authors, journal, year, and abstract. Ask before large batch
fetches; stay within NCBI rate limits.
- Refine with focused terms, MeSH-style phrasing, or a date/keyword narrow
when results are too broad.
- Report each kept article with title, PMID, journal/year, one-line
relevance, and the URL.
- Provenance & review: record queries, tool, and cited PMIDs with
$science-provenance; check claims with $science-review before presenting.
Boundaries
- Cite only PMIDs actually returned by the tool; never invent PMIDs, journals,
or abstracts. Verify the abstract before quoting it.
- Search results are references, not clinical advice; do not infer diagnosis,
dosing, or treatment from them.
- Prefer primary studies and systematic reviews over secondary summaries when the
user needs evidence strength.