National Cancer Institute (cancer.gov) MCP server for AI agents

Let Claude, ChatGPT or your own agent read National Cancer Institute (cancer.gov): 2 typed functions through one MCP server. No National Cancer Institute (cancer.gov) API key needed.

Functions
2
Sign-in
Not needed
Works with
MCP, HTTP, Node, Python

Connect

Just National Cancer Institute (cancer.gov)

An MCP server for National Cancer Institute (cancer.gov) alone, with one tool per function. Paste it into Claude, ChatGPT, Cursor or any MCP client as a remote server.

https://api.bowmark.ai/mcp/site/cancer
Every site

One MCP server for National Cancer Institute (cancer.gov) and every other site Bowmark covers. Your agent looks up what it needs and calls it.

https://api.bowmark.ai/mcp
Set it up in your client

Both need a Bowmark API key. The quickstart gets you one and a first result in a few minutes.

Functions

Open a function to see its arguments, what it returns and how to call it.

Read 2
  • findCancerCentersReturns the NCI-Designated Cancer Centers.

    Arguments

    NameTypeRequired
    args{ state?: string }No

    Returns cancerCentersResult

    type cancerCenterDesignation =
      | "Comprehensive Cancer Center"
      | "Clinical Cancer Center"
      | "Basic Laboratory Cancer Center";
    
    interface cancerCenterRow {
      name: string;
      /** The center's own detail page on cancer.gov. */
      url: string;
      /** NCI's own state grouping, the authoritative field to filter on. */
      state: string;
      /** The "City, State" line the page publishes, verbatim. */
      location: string;
      /** A second location aside the page publishes for a few centers, e.g.
       * "(in addition to facilities in Florida and Minnesota)", present on a
       * center that operates comprehensive facilities in more than one state and
       * is cross-listed under each. */
      locationNote?: string;
      /** The parent university or health system, when distinct from the center's
       * own name. Absent for freestanding centers (their own name IS the
       * institution). */
      hostInstitution?: string;
      designation: cancerCenterDesignation;
    }
    
    interface cancerCentersResult {
      /** Every center when no `state` filter is given; only the matching ones
       * otherwise. */
      total: number;
      centers: cancerCenterRow[];
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/cancer/findCancerCenters

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.cancer.findCancerCenters();
  • getCancerInfoSummaryReturns NCI's own PDQ information summary for one cancer type or topic.

    Arguments

    NameTypeRequired
    args{ topic: string; audience?: "patient" | "healthProfessional" }Yes

    Returns cancerInfoSummaryResult

    type cancerInfoAudience = "patient" | "healthProfessional";
    
    interface cancerInfoSection {
      /** The PDQ document's own heading, verbatim, e.g. "Stage Information for
       * Breast Cancer". */
      heading: string;
      /** The section's own text, whitespace-normalized. */
      text: string;
    }
    
    interface cancerInfoSummaryResult {
      topic: string;
      audience: cancerInfoAudience;
      /** The PDQ document's own title, verbatim. */
      title: string;
      /** The summary's own page on cancer.gov. */
      url: string;
      /** ISO 8601, the PDQ Editorial Board's own "Updated" date for this exact
       * document. PDQ is a living, continuously revised document; there is no
       * honest way to use this text without it. */
      revisionDate: string;
      sections: cancerInfoSection[];
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/cancer/getCancerInfoSummary

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.cancer.getCancerInfoSummary({ topic: "prostate" });

Use it from code