SNOMED Codes
Search and lookup SNOMED CT clinical concepts via the @sequoiaport/codes package.
When to Apply
- User asks for a SNOMED concept or concept ID
- User needs standardized clinical terminology
- User wants to map a clinical condition to SNOMED CT
- User mentions clinical ontology or concept hierarchies
Getting an API Key
First check if SEQUOIA_CODES_API_KEY is already set in the environment. If it is, skip key generation entirely and proceed to usage. Only if no key exists, see AGENT_PROMPT.md for how your agent can obtain one via the agent auth flow.
MCP Server
Once you have an API key, you can set up the MCP server for direct tool access from Claude Code, Cursor, or VS Code. See MCP_SETUP.md.
How It Works
import { SequoiaCodesClient } from "@sequoiaport/codes";
const client = new SequoiaCodesClient({
apiKey: process.env.SEQUOIA_CODES_API_KEY!,
});
Available Operations
client.snomed.searchCode({ query, limit? })
Hybrid search for SNOMED CT concepts.
const results = await client.snomed.searchCode({
query: "congestive heart failure",
limit: 10,
});
// results.results → [{ concept_id, fsn, preferred_term, semantic_tag, similarity, icd10_codes, ... }]
client.snomed.identifyCode({ code })
Look up a specific SNOMED concept by concept ID.
const result = await client.snomed.identifyCode({ code: "73211009" });
// result.concept → { concept_id: "73211009", preferred_term: "Diabetes mellitus", semantic_tag: "disorder", ... }
REST API (curl)
# Search SNOMED concepts
curl -X GET "https://api.sequoiacodes.com/v1/snomed/searchCode?query=heart%20failure&limit=10" \
-H "Authorization: Bearer $SEQUOIA_CODES_API_KEY"
# Lookup concept by ID
curl -X GET "https://api.sequoiacodes.com/v1/snomed/identifyCode?code=73211009" \
-H "Authorization: Bearer $SEQUOIA_CODES_API_KEY"
Response Fields
Each SNOMED concept includes:
concept_id- SNOMED CT concept IDfsn- Fully Specified Namepreferred_term- Preferred display termsemantic_tag- Category (disorder, finding, procedure, etc.)definition_status- Primitive or fully definedactive- Whether concept is activesimilarity- Search relevance score (0-1)icd10_codes- Mapped ICD-10 codes (when available)