pubmed-mcp-server

Pubmed Fetch Citation Metrics

pubmed_fetch_citation_metrics

Read-only

Fetch NIH iCite citation metrics for PubMed articles: the Relative Citation Ratio (RCR), NIH percentile, citation count, citations per year, expected and field citation rates, the Approximate Potential to Translate (APT), the human, animal, and molecular/cellular shares of each article's MeSH terms, and how many clinical articles cite it. RCR is a field- and year-normalized citation rate benchmarked to NIH-funded papers, whose median RCR is 1.0; iCite computes none for the most recent papers and marks recent ones provisional. iCite does not flag retractions — a retracted article keeps its metrics — so check pubmed_fetch_articles for retraction notices before relying on a record. Up to 1000 PMIDs per call; the first 200 distinct PMIDs are fetched and the rest returned in overLimit. PMIDs iCite has no record for are listed in unavailablePmids.

Input Schema

Name	Required	Description	Default
`pmids`	Yes	PubMed IDs to look up (e.g. ["23456789"]). The first 200 distinct PMIDs are fetched; the rest are returned in `overLimit`. Leading zeros are ignored, so "00023456789" and "23456789" name one article.

Output Schema

Name	Required	Description
`error`	No	Present when the call failed. Absent on success.
`notice`	No	Names the PMIDs iCite had no record for, records not yet scored, records with no RCR or a provisional one, and PMIDs past the per-call limit with how to fetch them. Absent when every record came back final and complete.
`records`	No	One record per fetched PMID iCite knows, in request order.
`overLimit`	No	PMIDs the call named past the per-call fetch limit. Present only when the call named more distinct PMIDs than one call fetches.
`unavailablePmids`	No	Requested PMIDs within the per-call limit that NIH iCite has no record for, in request order, under the spelling sent. iCite omits a PMID it has no record for without an error, so the absence says nothing about whether PubMed has the article; pubmed_fetch_articles confirms that.

Schema Changelog

Changes observed during successful MCP inspections.

Changed2 schema fields changedNov 3, 2026

changedOutput schema / properties / error / properties / data / properties / reason / description

Previous value: -"Machine-readable failure mode. Declared by this tool: `icite_unreachable`: NIH iCite failed on every attempt the retry budget allowed — a network failure or an HTTP 5xx. `icite_invalid_response`: NIH iCite returned a body that is not its JSON record list on the last attempt the retry budget allowed, or a record for a PMID the call did not request. Other values are possible when a failure originates below the handler."New value: +"Machine-readable failure mode. Declared by this tool: `icite_unreachable`: NIH iCite failed on every attempt the retry budget allowed — a network failure or an HTTP 5xx. `icite_invalid_response`: NIH iCite returned a body that is not its JSON record list on the last attempt the retry budget allowed. `icite_unrequested_record`: NIH iCite answered with a record for a PMID the call did not request, or one whose PMID is not a PMID — the answer it gives this PMID list every time. Other values are possible when a failure originates below the handler."

changedOutput schema / properties / error / properties / data / properties / reason / examples

Previous value: -[
-  "icite_unreachable",
-  "icite_invalid_response"
-]New value: +[
+  "icite_unreachable",
+  "icite_invalid_response",
+  "icite_unrequested_record"
+]

AddedOct 25, 2026

TDQS

A4.4/5.0

Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

The description discloses behavior beyond the annotations: iCite omits metrics for recent papers, marks some provisional, does not flag retractions, and returns over-limit PMIDs in `overLimit` and missing records in `unavailablePmids`. These are important operational caveats for correct use.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is front-loaded with the tool's purpose and metrics, followed by behavioral caveats and limits. It is dense but well structured; the metric enumeration is long yet useful for understanding what is returned.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the output schema, annotations, and single required parameter, the description is complete. It covers the metrics returned, recent-paper limitations, retraction caveats, and over-limit behavior, leaving no critical gaps for an agent invoking the tool.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so the single `pmids` parameter is already documented. The description reinforces the 1000-PMID limit and first-200-distinct behavior, but adds little semantic meaning beyond what the schema already provides.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description states a specific verb and resource: fetch NIH iCite citation metrics for PubMed articles. It enumerates the exact metrics returned (RCR, percentile, citation count, APT, MeSH shares, citing clinical articles), making the tool distinguishable from general article-fetching siblings.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

It gives clear usage context, including the retraction caveat and directs the agent to pubmed_fetch_articles for retraction notices before relying on metrics. It does not explicitly compare against other citation-related siblings such as pubmed_lookup_citation or pubmed_format_citations, but the intended use case is clear.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

All pubmed-mcp-server tools Inspect

Try in Browser

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.