Claude SNOMED CT Terminology connectorWhat Claude can look up in SNOMED CT, straight from SNOMED International.
The Claude SNOMED CT Terminology connector exposes 15 tools, all of them for reading. 0 write, 0 undocumented. They look up, validate, compare and expand clinical concepts on SNOMED International's server. Below: what each returns and where the server draws lines.
Verified Trustpilot reviews · AI, automation & growth agency
What changes when Claude can query SNOMED CT
SNOMED CT is the clinical terminology used in electronic health records across more than 80 countries. With this connector, Claude queries it directly through a terminology server run by SNOMED International, the owner of SNOMED CT. You ask about a code or a condition in plain words and get the answer from the live terminology.
A code checked before it is used. snomed_lookup returns a concept's names, parents and defining attributes, and snomed_validate_code says whether it is still active.
A clean concept set for a rule or a report. snomed_expand runs queries in Expression Constraint Language (ECL, SNOMED's query syntax) to pull every subtype or every concept matching an attribute in one list.
The hierarchy made visible. snomed_get_ancestors, snomed_get_children and snomed_get_descendants place a concept in the IS-A tree, and snomed_subsumes tests whether one concept falls under another.
What it will not do: the publisher states every tool is read-only and the connector holds no query content between requests. There is no account here, no patient record and nothing to edit. Nothing runs by itself either. For flows that react to events, the Integrations hub covers automation tools, a different job, and the Claude connectors page lists the rest.
Five words before you connect
The vocabulary around this connector, in one minute.
- Connector
- A link you set up once between Claude and an outside service, so Claude can use that service while it answers you.
- Tool
- One named action the connector gives Claude. Claude picks the tools it needs as it answers, and the directory sheet lists them all.
- Approval
- The confirmation Claude waits for before it goes ahead with something that changes data, shown in the conversation when it matters.
- Directory
- The official catalog where each connector has a sheet with its publisher, its purpose and the list of tools it exposes.
- MCP
- The shared standard connectors are built on: it is what lets an assistant like Claude query an outside service such as this terminology server.
Connect the Claude SNOMED CT Terminology connector in three steps
- 01
Find the connector in Claude
Open Claude's settings, go to Customize then Connectors, and look up SNOMED CT Terminology in the list. On a Team or Enterprise workspace, an Owner or Primary Owner has to enable it for the organization first.
- 02
Start the connection
Click Connect on its row. SNOMED International describes the server as a stateless proxy that keeps no query content or personal data between requests; no source describes a personal account to sign in with.
- 03
Ask about a concept
Open a conversation and ask about a code or a clinical term. If a link ever breaks, Disconnect and connect again from the same list. Claude then picks the lookup or search tool that fits.
The 15 tools, all on the reading side
SNOMED CT Terminology gives Claude 15 tools: 15 that read your account, 0 that change something in it.
SNOMED International's server marks all 15 tools as read only. Tool names stay exactly as Claude shows them.
- 15 read
What Claude reads (15)
15 toolsFifteen tools that look things up in the terminology and leave it as it is.
fhir_metadata
Returns a readable summary of the FHIR capability statement published by a terminology's backend, with the raw statement attached by default unless Claude asks to leave it out.
list_terminologies
Lists the SNOMED CT editions and branches available on the connected server, so the right terminology key is known before any edition-specific query.
server_capabilities
Reports the backend classification and capability flags for a given terminology, on the server's default one unless another is named.
server_health
Checks whether the backend is reachable and returns its basic capability flags, for the default terminology or one you specify.
snomed_expand
Expands a SNOMED CT value set through FHIR, which the server presents as its primary tool for retrieving concept sets with Expression Constraint Language queries.
snomed_get_ancestors
Climbs the IS-A hierarchy from a concept to its parents, grandparents and beyond, or stops at the direct parents if asked, to show where the concept is classified.
snomed_get_children
Returns the direct IS-A children of a concept, one level down, which the server describes as lighter than ECL for simple browsing of a subhierarchy.
snomed_get_descendants
Gathers every descendant of a concept, children, grandchildren and further down, returned as a hierarchy that can be paged.
snomed_lookup
Fetches the full detail of a known concept ID: fully specified name, synonyms, parents and defining attributes, and whether it is active.
snomed_subsumes
Tests subsumption between two SNOMED CT codes through the FHIR operation of the same name, that is whether one concept falls under the other.
snomed_validate_code
Confirms whether a concept ID exists and is currently active, and gives back its display name when it is.
snowstorm_get_concept_native
Fetches the detail of a concept by its ID through Snowstorm's own interface rather than the FHIR one, on the default terminology unless another is named.
snowstorm_list_codesystems
Lists the Snowstorm code systems the server knows, each one with a short summary of its latest version, on the default terminology unless another one is named.
snowstorm_list_versions
Lists the versions available for one Snowstorm code system, identified by its short name, so you can see which releases exist.
snowstorm_search_concepts
Searches concepts by term through Snowstorm's own search, outside FHIR.
What Claude checks with you first
By default, Claude stops and asks before each action it takes for someone in an outside service. The prompt appears in the conversation when it matters.
On Team and Enterprise workspaces, owners decide whether a member may let some actions through without a prompt each time. They can also limit what a connector may do for the whole organization, and that setting binds everyone. Here every listed tool reads a public terminology, and no source describes a confirmation for any of them. The server does ask Claude to check with you before fetching very large result sets.
Which plans include it
None of the 819 sheets in the official directory shows plan availability, so the answer for this connector is published nowhere.
The published rule is general: remote connectors are open to all users on Claude, Cowork, Claude Desktop and mobile. On Team and Enterprise, an Owner or Primary Owner opens a connector for the organization before members can use it. For the current state of this connector, check its sheet in the official directory.
Where this connector stops
A connector is not an automation. Claude calls these tools while it answers you, so no code gets checked when a record changes.
Everything specific here comes from SNOMED International, through the directory sheet and the tool descriptions on its server: no Claude help article covers this connector. A list showing only reading tools is a floor, not a promise. The verified badge is not a security audit either, and Anthropic states on every sheet that it neither picks the tools a publisher exposes nor guarantees how they behave.
Need help connecting SNOMED CT Terminology to Claude?
A person reads every message.