pubmed-mcp-server

Pubmed Spell Check

pubmed_spell_check

Read-only

Spell-check a PubMed search query against NCBI ESpell and get the corrected query back. Use after a zero-hit or thin pubmed_search_articles result, or when a drug, gene, disease, or author name may be misspelled. Corrections are applied in place, so field tags, operators, spacing, and gene symbols stay as written (alzhiemer diseese treatmnt outcomse → alzheimer disease treatment outcomes). The value of an indexed field — MeSH, publication type, journal, author, language, subset, and the like (aged[mh], free full text[sb], zhang y[au]) — is kept as written, misspelled or not, since the ESpell dictionary is not the index that field searches; free-text fields such as [tiab], [ti], and [tw] are corrected. Words in all capitals, with digits or non-ASCII letters, or differing only in case are never changed, nor is a single letter swapped for another (q fever, the initial in smith q), words after a number or symbol ESpell skips are left as written, and hasSuggestion is false when no correction was applied. Re-run the search with corrected.

Input Schema

Name	Required	Description	Default
`query`	Yes	PubMed search query to spell-check. Must carry a term: a value of only whitespace or invisible characters, such as a zero-width space, is rejected rather than sent to ESpell.

Output Schema

Name	Required	Description
`error`	No	Present when the call failed. Absent on success.
`original`	No	Original query
`corrected`	No	The query with ESpell's accepted corrections applied in place; field tags, the values of indexed fields such as `aged[mh]` and `zhang y[au]`, operators, spacing, words in capitals or with digits, and a single letter ESpell would swap for another are kept as written. Same as `original` when `hasSuggestion` is false.
`hasSuggestion`	No	Whether at least one correction was applied to `corrected`

Schema Changelog

Changes observed during successful MCP inspections.

Changed1 schema field changedNov 3, 2026

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

Previous value: -"Machine-readable failure mode. Declared by this tool: `queue_full`: The local NCBI request queue shed the call — the queue is full, or the call cannot start before its total deadline (for example behind the cooldown that follows an NCBI 429). `ncbi_unreachable`: NCBI E-utilities failed on every attempt the retry budget allowed — retries ran out, or the next backoff would overrun the total deadline. `ncbi_rate_limited`: NCBI answered HTTP 429 (too many requests) and the call stopped on it — retries ran out, the next backoff would overrun the total deadline, or the Retry-After NCBI named outlasts the time left or the 30-second backoff cap. `ncbi_deadline_exceeded`: The total NCBI request deadline expired before NCBI answered successfully — mid-request, while queued, or during a retry backoff. `ncbi_invalid_response`: NCBI returned a body that could not be parsed (invalid XML/JSON). `ncbi_resource_not_found`: NCBI returned a structured \"not found\" error for the requested ID(s). `blank_query`: The query contains no search term: nothing is left once whitespace and invisible characters such as a zero-width space are disregarded. pubmed_search_articles and pubmed_europepmc_search first decode HTML entities and also disregard markup and parentheses, so `()`, `<b></b>`, and `&nbsp;` hold no term there; pubmed_search_articles also disregards bracketed field tags such as `[pdat]`. Other values are possible when a failure originates below the handler."New value: +"Machine-readable failure mode. Declared by this tool: `queue_full`: The local NCBI request queue shed the call — the queue is full, or the call cannot start before its total deadline (for example behind the cooldown that follows an NCBI 429). `ncbi_unreachable`: NCBI E-utilities failed on every attempt the retry budget allowed — retries ran out, or the next backoff would overrun the total deadline. `ncbi_rate_limited`: NCBI answered HTTP 429 (too many requests) and the call stopped on it — retries ran out, the next backoff would overrun the total deadline, or the Retry-After NCBI named outlasts the time left or the 30-second backoff cap. `ncbi_deadline_exceeded`: The total NCBI request deadline expired before NCBI answered successfully — mid-request, while queued, or during a retry backoff. `ncbi_invalid_response`: NCBI returned a body that could not be parsed (invalid XML/JSON). `ncbi_resource_not_found`: NCBI returned a structured \"not found\" error for the requested ID(s). `blank_query`: The query contains no search term: nothing is left once whitespace and invisible characters such as a zero-width space are disregarded. pubmed_search_articles and pubmed_europepmc_search first decode HTML entities and also disregard markup and parentheses, so `()`, `<b></b>`, and `&nbsp;` hold no term there; pubmed_search_articles also disregards bracketed field tags such as `[pdat]`. On pubmed_search_articles an exactly-empty `query` is not blank: it means no query, and the search runs on its filters. Other values are possible when a failure originates below the handler."

Changed1 schema field changedOct 25, 2026

changedOutput schema / properties / corrected / description

Previous value: -"The query with ESpell's accepted corrections applied in place; field tags, operators, spacing, words in capitals or with digits, and a single letter ESpell would swap for another are kept as written. Same as `original` when `hasSuggestion` is false."New value: +"The query with ESpell's accepted corrections applied in place; field tags, the values of indexed fields such as `aged[mh]` and `zhang y[au]`, operators, spacing, words in capitals or with digits, and a single letter ESpell would swap for another are kept as written. Same as `original` when `hasSuggestion` is false."

Changed1 schema field changedOct 23, 2026

changedOutput schema / properties / corrected / description

Previous value: -"The query with ESpell's accepted corrections applied in place; field tags, operators, spacing, and words in capitals or with digits are kept as written. Same as `original` when `hasSuggestion` is false."New value: +"The query with ESpell's accepted corrections applied in place; field tags, operators, spacing, words in capitals or with digits, and a single letter ESpell would swap for another are kept as written. Same as `original` when `hasSuggestion` is false."

Changed2 schema fields changedOct 12, 2026

changedOutput schema / properties / corrected / description

Previous value: -"Corrected query (same as original if no suggestion)"New value: +"The query with ESpell's accepted corrections applied in place; field tags, operators, spacing, and words in capitals or with digits are kept as written. Same as `original` when `hasSuggestion` is false."

changedOutput schema / properties / hasSuggestion / description

Previous value: -"Whether NCBI suggested a correction"New value: +"Whether at least one correction was applied to `corrected`"

Changed3 schema fields changedSep 28, 2026

changedInput schema / properties / query / description

Previous value: -"PubMed search query to spell-check. Must carry a term: a blank or whitespace-only value is rejected rather than sent to ESpell."New value: +"PubMed search query to spell-check. Must carry a term: a value of only whitespace or invisible characters, such as a zero-width space, is rejected rather than sent to ESpell."

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

Previous value: -"Machine-readable failure mode. Declared by this tool: `queue_full`: The local NCBI request queue shed the call — the queue is full, or the call cannot start before its total deadline (for example behind the cooldown that follows an NCBI 429). `ncbi_unreachable`: NCBI E-utilities is unreachable after all retry attempts. `ncbi_deadline_exceeded`: Total request deadline expired before NCBI returned a response. `ncbi_invalid_response`: NCBI returned a body that could not be parsed (invalid XML/JSON). `ncbi_resource_not_found`: NCBI returned a structured \"not found\" error for the requested ID(s). `blank_query`: The query holds no search term once whitespace, and anything the tool strips before searching, are removed — so NCBI would receive a blank term. pubmed_search_articles strips markup, bracketed field tags, and parentheses, so a bare field tag such as `[pdat]` or empty parentheses `()` count as blank there. Other values are possible when a failure originates below the handler."New value: +"Machine-readable failure mode. Declared by this tool: `queue_full`: The local NCBI request queue shed the call — the queue is full, or the call cannot start before its total deadline (for example behind the cooldown that follows an NCBI 429). `ncbi_unreachable`: NCBI E-utilities failed on every attempt the retry budget allowed — retries ran out, or the next backoff would overrun the total deadline. `ncbi_rate_limited`: NCBI answered HTTP 429 (too many requests) and the call stopped on it — retries ran out, the next backoff would overrun the total deadline, or the Retry-After NCBI named outlasts the time left or the 30-second backoff cap. `ncbi_deadline_exceeded`: The total NCBI request deadline expired before NCBI answered successfully — mid-request, while queued, or during a retry backoff. `ncbi_invalid_response`: NCBI returned a body that could not be parsed (invalid XML/JSON). `ncbi_resource_not_found`: NCBI returned a structured \"not found\" error for the requested ID(s). `blank_query`: The query contains no search term: nothing is left once whitespace and invisible characters such as a zero-width space are disregarded. pubmed_search_articles and pubmed_europepmc_search first decode HTML entities and also disregard markup and parentheses, so `()`, `<b></b>`, and `&nbsp;` hold no term there; pubmed_search_articles also disregards bracketed field tags such as `[pdat]`. Other values are possible when a failure originates below the handler."

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

Previous value: -[
-  "queue_full",
-  "ncbi_unreachable",
-  "ncbi_deadline_exceeded",
-  "ncbi_invalid_response",
-  "ncbi_resource_not_found",
-  "blank_query"
-]New value: +[
+  "queue_full",
+  "ncbi_unreachable",
+  "ncbi_rate_limited",
+  "ncbi_deadline_exceeded",
+  "ncbi_invalid_response",
+  "ncbi_resource_not_found",
+  "blank_query"
+]

Changed1 schema field changedSep 23, 2026

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

Previous value: -"Machine-readable failure mode. Declared by this tool: `queue_full`: The local NCBI request queue shed the call — the queue is full, or the call cannot start before its total deadline (for example behind the cooldown that follows an NCBI 429). `ncbi_unreachable`: NCBI E-utilities is unreachable after all retry attempts. `ncbi_deadline_exceeded`: Total request deadline expired before NCBI returned a response. `ncbi_invalid_response`: NCBI returned a body that could not be parsed (invalid XML/JSON). `ncbi_resource_not_found`: NCBI returned a structured \"not found\" error for the requested ID(s). `blank_query`: The query holds no search term once whitespace, and any markup the tool strips first, are removed — so NCBI would receive a blank term. Other values are possible when a failure originates below the handler."New value: +"Machine-readable failure mode. Declared by this tool: `queue_full`: The local NCBI request queue shed the call — the queue is full, or the call cannot start before its total deadline (for example behind the cooldown that follows an NCBI 429). `ncbi_unreachable`: NCBI E-utilities is unreachable after all retry attempts. `ncbi_deadline_exceeded`: Total request deadline expired before NCBI returned a response. `ncbi_invalid_response`: NCBI returned a body that could not be parsed (invalid XML/JSON). `ncbi_resource_not_found`: NCBI returned a structured \"not found\" error for the requested ID(s). `blank_query`: The query holds no search term once whitespace, and anything the tool strips before searching, are removed — so NCBI would receive a blank term. pubmed_search_articles strips markup, bracketed field tags, and parentheses, so a bare field tag such as `[pdat]` or empty parentheses `()` count as blank there. Other values are possible when a failure originates below the handler."

Changed1 schema field changedSep 23, 2026

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

Previous value: -"Machine-readable failure mode. Declared by this tool: `queue_full`: Local NCBI request queue is at capacity. `ncbi_unreachable`: NCBI E-utilities is unreachable after all retry attempts. `ncbi_deadline_exceeded`: Total request deadline expired before NCBI returned a response. `ncbi_invalid_response`: NCBI returned a body that could not be parsed (invalid XML/JSON). `ncbi_resource_not_found`: NCBI returned a structured \"not found\" error for the requested ID(s). `blank_query`: The query holds no search term once whitespace, and any markup the tool strips first, are removed — so NCBI would receive a blank term. Other values are possible when a failure originates below the handler."New value: +"Machine-readable failure mode. Declared by this tool: `queue_full`: The local NCBI request queue shed the call — the queue is full, or the call cannot start before its total deadline (for example behind the cooldown that follows an NCBI 429). `ncbi_unreachable`: NCBI E-utilities is unreachable after all retry attempts. `ncbi_deadline_exceeded`: Total request deadline expired before NCBI returned a response. `ncbi_invalid_response`: NCBI returned a body that could not be parsed (invalid XML/JSON). `ncbi_resource_not_found`: NCBI returned a structured \"not found\" error for the requested ID(s). `blank_query`: The query holds no search term once whitespace, and any markup the tool strips first, are removed — so NCBI would receive a blank term. Other values are possible when a failure originates below the handler."

Changed3 schema fields changedSep 11, 2026

changedInput schema / properties / query / description

Previous value: -"PubMed search query to spell-check"New value: +"PubMed search query to spell-check. Must carry a term: a blank or whitespace-only value is rejected rather than sent to ESpell."

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

Previous value: -"Machine-readable failure mode. Declared by this tool: `queue_full`: Local NCBI request queue is at capacity. `ncbi_unreachable`: NCBI E-utilities is unreachable after all retry attempts. `ncbi_deadline_exceeded`: Total request deadline expired before NCBI returned a response. `ncbi_invalid_response`: NCBI returned a body that could not be parsed (invalid XML/JSON). `ncbi_resource_not_found`: NCBI returned a structured \"not found\" error for the requested ID(s). Other values are possible when a failure originates below the handler."New value: +"Machine-readable failure mode. Declared by this tool: `queue_full`: Local NCBI request queue is at capacity. `ncbi_unreachable`: NCBI E-utilities is unreachable after all retry attempts. `ncbi_deadline_exceeded`: Total request deadline expired before NCBI returned a response. `ncbi_invalid_response`: NCBI returned a body that could not be parsed (invalid XML/JSON). `ncbi_resource_not_found`: NCBI returned a structured \"not found\" error for the requested ID(s). `blank_query`: The query holds no search term once whitespace, and any markup the tool strips first, are removed — so NCBI would receive a blank term. Other values are possible when a failure originates below the handler."

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

Previous value: -[
-  "queue_full",
-  "ncbi_unreachable",
-  "ncbi_deadline_exceeded",
-  "ncbi_invalid_response",
-  "ncbi_resource_not_found"
-]New value: +[
+  "queue_full",
+  "ncbi_unreachable",
+  "ncbi_deadline_exceeded",
+  "ncbi_invalid_response",
+  "ncbi_resource_not_found",
+  "blank_query"
+]

Changed6 schema fields changedAug 21, 2026

changedInput schema / $schema

Previous value: -"http://json-schema.org/draft-07/schema#"New value: +"https://json-schema.org/draft/2020-12/schema"

addedInput schema / additionalProperties
```
Added value: +false
```

changedOutput schema / $schema

Previous value: -"http://json-schema.org/draft-07/schema#"New value: +"https://json-schema.org/draft/2020-12/schema"

addedOutput schema / anyOf

Added value: +[
+  {
+    "not": {
+      "required": [
+        "error"
+      ]
+    },
+    "required": [
+      "original",
+      "corrected",
+      "hasSuggestion"
+    ]
+  },
+  {
+    "required": [
+      "error"
+    ]
+  }
+]

addedOutput schema / properties / error

Added value: +{
+  "additionalProperties": {},
+  "description": "Present when the call failed. Absent on success.",
+  "properties": {
+    "code": {
+      "description": "JSON-RPC error code for this failure.",
+      "maximum": 9007199254740991,
+      "minimum": -9007199254740991,
+      "type": "integer"
+    },
+    "data": {
+      "additionalProperties": {},
+      "properties": {
+        "reason": {
+          "description": "Machine-readable failure mode. Declared by this tool: `queue_full`: Local NCBI request queue is at capacity. `ncbi_unreachable`: NCBI E-utilities is unreachable after all retry attempts. `ncbi_deadline_exceeded`: Total request deadline expired before NCBI returned a response. `ncbi_invalid_response`: NCBI returned a body that could not be parsed (invalid XML/JSON). `ncbi_resource_not_found`: NCBI returned a structured \"not found\" error for the requested ID(s). Other values are possible when a failure originates below the handler.",
+          "examples": [
+            "queue_full",
+            "ncbi_unreachable",
+            "ncbi_deadline_exceeded",
+            "ncbi_invalid_response",
+            "ncbi_resource_not_found"
+          ],
+          "type": "string"
+        },
+        "recovery": {
+          "additionalProperties": {},
+          "description": "Actionable next step for the caller.",
+          "properties": {
+            "hint": {
+              "type": "string"
+            }
+          },
+          "required": [
+            "hint"
+          ],
+          "type": "object"
+        },
+        "retryable": {
+          "description": "Whether retrying may succeed.",
+          "type": "boolean"
+        }
+      },
+      "type": "object"
+    },
+    "message": {
+      "description": "Human-readable description of what went wrong.",
+      "type": "string"
+    }
+  },
+  "required": [
+    "code",
+    "message"
+  ],
+  "type": "object"
+}

removedOutput schema / required

Removed value: -[
-  "original",
-  "corrected",
-  "hasSuggestion"
-]

First observedMar 23, 2026

TDQS

A4.6/5.0

Behavior5/5

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

Annotations only say readOnly and openWorld; the description goes far beyond by explaining that corrections are applied in place, that indexed field values (`[mh]`, `[sb]`, `[au]`) are preserved while free-text fields (`[tiab]`, `[ti]`, `[tw]`) are corrected, and which token classes are never touched. It even documents the `hasSuggestion` flag, which is unusual and useful given ESpell's quirky dictionary behavior.

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?

Front-loaded with purpose and trigger before the detailed correction rules, and the worked example (`alzhiemer diseese treatmnt outcomse` → ...) anchors the abstract rules. The mid-paragraph enumeration of exception cases is dense and slightly longer than needed, but every clause conveys a real edge case.

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?

With an output schema present, the description doesn't need to restate return shape, and it still flags the `hasSuggestion` signal. For a single-parameter tool, the description covers validity, correction semantics, exceptions, and next step — nothing an agent needs is missing.

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?

There is one fully documented parameter (100% schema coverage), so the schema already carries the baseline. The description adds contextual meaning — that the query may contain operators and field tags that survive correction — but does not spell out additional format rules for `query` beyond the schema's non-blank constraint.

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?

States a specific verb and resource ('Spell-check a PubMed search query') plus the backing service (NCBI ESpell) and the return value (corrected query). It also names the sibling `pubmed_search_articles` as the upstream step, so an agent can place it in the workflow without opening a schema.

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

Usage Guidelines5/5

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

Gives concrete triggers — a zero-hit or thin `pubmed_search_articles` result, or suspicion that a drug, gene, disease, or author name is misspelled — and closes the loop with 'Re-run the search with `corrected`.' When-to-use and the follow-up action are both explicit.

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.

Pubmed Spell Check

Input Schema

Output Schema

Schema Changelog

TDQS

Glama MCP Gateway

Resources