pubmed-clinical-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| NCBI_EMAIL | No | Email address recommended for NCBI E-utilities usage. | |
| NCBI_API_KEY | No | Optional NCBI API key to raise rate limits. | |
| UNPAYWALL_EMAIL | No | Email address required for Unpaywall checks. | |
| PUBMED_MCP_CACHE_PATH | No | Path to a local JSON cache file. Optional. |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| build_clinical_queryB | Convert a natural-language clinical question into a PubMed-ready query. |
| search_pubmedB | Search PubMed and return PMIDs plus a lightweight metadata summary. |
| fetch_pubmed_articlesC | Fetch PubMed title, abstract, authors, journal, year, DOI, MeSH, and publication types. |
| rank_articlesA | Rank article summaries against a clinical question and explain relevance. |
| search_and_fetch_pubmedC | Build a clinical PubMed query, search PubMed, fetch articles, and return ranked summaries. |
| fetch_pmc_full_textB | Fetch legal PubMed Central full-text sections when a PMCID is available. |
| check_full_text_availabilityC | Check PMC and Unpaywall for legal full-text availability. |
| find_related_articlesC | Find related PubMed articles through NCBI elink. |
| extract_picoB | Extract a rough PICO structure from an abstract or article passage. |
| summarize_evidenceC | Summarize article evidence into a table, limitations, and citation-backed findings. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 10 tools
Each tool has a clearly distinct purpose: building queries, checking availability, extracting PICO, fetching full text, fetching articles, finding related articles, ranking, searching (with and without full pipeline), and summarizing. No ambiguity.
All tool names follow a consistent verb_noun pattern in snake_case (e.g., build_clinical_query, fetch_pubmed_articles, summarize_evidence). No deviations or mixed conventions.
10 tools is well-scoped for a clinical PubMed assistant. Each tool serves a distinct step in the research workflow without unnecessary overlap or missing critical functionality.
The tool set covers the full clinical evidence retrieval workflow: natural language query conversion, search, article fetch, full text retrieval, PICO extraction, ranking, summarization, and related article discovery. No obvious gaps.