Wikipedia MCP server for AI agents

Let Claude, ChatGPT or your own agent read and act on Wikipedia: 40 typed functions through one MCP server. No Wikipedia API key needed.

Functions
40
Sign-in
For account actions
Works with
MCP, HTTP, Node, Python

Connect

Just Wikipedia

An MCP server for Wikipedia 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/wikipedia
Every site

One MCP server for Wikipedia 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 that need a login use the person's own account: how sign-in works.

Functions

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

Read 37
  • compareRevisionsThe diff between two revisions of an article.

    Arguments

    NameTypeRequired
    fromRevisionIdnumberYes
    toRevisionIdnumberYes
    options{ lang?: string }No

    Returns WikipediaRevisionComparison

    interface WikipediaRevisionDiffLine {
      type: number;  // MediaWiki's own line-diff code: 0 unchanged (context), 3 changed (see highlightRanges); 1/2 (pure add/remove) pass through unchanged
      lineNumber: number;
      text: string;
      offset: { from: number | null; to: number | null };
      highlightRanges?: { start: number; length: number; type: number }[];
    }
    
    interface WikipediaRevisionComparison {
      from: number;
      to: number;
      diff: WikipediaRevisionDiffLine[];
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/compareRevisions

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.compareRevisions(1369844273, 1374050869);
  • getArticleThe whole article as an agent wants to read it: title, Wikipedia's own short description, the full body as plain text with the markup gone, the…

    Arguments

    NameTypeRequired
    titleOrUrlstringYes
    options{ lang?: string }No

    Returns WikipediaArticle

    interface WikipediaArticle {
      id: number;
      title: string;
      url: string;
      description: string;
      extract: string;
      thumbnail?: {
        url: string;
        width: number;
        height: number;
      };
      coordinates?: {
        lat: number;
        lon: number;
      };
      revisionId: number;
      lastModified: string;
      length: number;
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/getArticle

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.getArticle("Espresso");
  • getArticleHtmlThe article's rendered HTML.

    Arguments

    NameTypeRequired
    titleOrUrlstringYes
    options{ lang?: string }No

    Returns WikipediaArticleHtml

    interface WikipediaArticleHtml {
      title: string;
      url: string;
      html: string;
      revisionId: number;
      lastModified: string;
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/getArticleHtml

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.getArticleHtml("Espresso");
  • getCurrentUserWho the caller is signed in as on Wikipedia.

    Arguments

    None.

    Returns WikipediaCurrentUser

    interface WikipediaCurrentUser {
      id: number;
      name: string;
      editCount: number;
      registrationTime: string;
      isBlocked: boolean;
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/getCurrentUser

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.getCurrentUser();
  • getFeaturedContentWikipedia's own front page for a given date, as data.

    Arguments

    NameTypeRequired
    options{ date?: string; lang?: string }No

    Returns { content: WikipediaFeaturedContent; warnings: string[] }

    interface WikipediaFeaturedArticle {
      title: string;
      description: string | null;
      extract: string;
      url: string;
      thumbnailUrl: string | null;
    }
    
    interface WikipediaMostReadArticle {
      title: string;
      url: string;
      views: number;
      rank: number;
    }
    
    interface WikipediaFeaturedImage {
      title: string;
      imageUrl: string;
      thumbnailUrl: string | null;
      artist: string | null;
      license: string | null;
      description: string | null;
    }
    
    interface WikipediaFeaturedNews {
      story: string;
      links: { title: string; url: string }[];
    }
    
    interface WikipediaFeaturedContent {
      date: string;
      featuredArticle: WikipediaFeaturedArticle | null;
      mostRead: WikipediaMostReadArticle[];
      image: WikipediaFeaturedImage | null;
      news: WikipediaFeaturedNews[];
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/getFeaturedContent

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.getFeaturedContent();
  • getImageOne media file's real details.

    Arguments

    NameTypeRequired
    fileTitlestringYes
    options{ lang?: string }No

    Returns WikipediaImage

    interface WikipediaImage {
      title: string;
      url: string;
      width: number;
      height: number;
      mime: string;
      descriptionUrl: string;
      license?: string;
      attribution?: string;
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/getImage

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.getImage("File:Espresso.jpg");
  • getInfoboxThe grey fact box at the top right of an article, as key/value pairs a caller can actually use.

    Arguments

    NameTypeRequired
    titleOrUrlstringYes
    options{ lang?: string }No

    Returns WikipediaInfobox

    interface WikipediaInfobox {
      type: string;
      /** Raw wikitext per field, as editors wrote it. */
      fields: Record<string, string>;
      /** The same fields as plain readable text (links, refs, comments and common templates resolved; empty values dropped). */
      facts: Record<string, string>;
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/getInfobox

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.getInfobox("Espresso");
  • getOnThisDayWhat happened on this calendar day in history, according to Wikipedia.

    Arguments

    NameTypeRequired
    options{ type?: string; date?: string; lang?: string }No

    Returns { onThisDay: WikipediaOnThisDay; warnings: string[] }

    interface WikipediaOnThisDayItem {
      year: string;
      text: string;
      pages: { title: string; url: string }[];
    }
    
    interface WikipediaOnThisDay {
      date: string;
      type: "all" | "births" | "deaths" | "events" | "holidays";
      items: WikipediaOnThisDayItem[];
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/getOnThisDay

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.getOnThisDay();
  • getPageviewsHow many people actually read an article, per day, over a date range.

    Arguments

    NameTypeRequired
    titleOrUrlstringYes
    options{ lang?: string; from?: string; to?: string }No

    Returns { pageviews: WikipediaPageviews[]; warnings: string[] }

    interface WikipediaPageviews {
      date: string;
      views: number;
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/getPageviews

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.getPageviews("Espresso");
  • getRandomArticleA genuinely random Wikipedia article.

    Arguments

    NameTypeRequired
    options{ lang?: string }No

    Returns { randomArticle: WikipediaRandomArticle; warnings: string[] }

    interface WikipediaRandomArticle {
      title: string;
      url: string;
      description: string;
      extract: string;
      thumbnail?: {
        url: string;
        width: number;
        height: number;
      };
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/getRandomArticle

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.getRandomArticle();
  • getRevisionOne specific revision of an article by id.

    Arguments

    NameTypeRequired
    revisionIdnumberYes
    options{ lang?: string }No

    Returns WikipediaRevisionDetail

    interface WikipediaRevisionDetail {
      id: number;
      page: { id: number; title: string };
      timestamp: string;
      user: string;
      comment: string;
      size: number;
      delta: number | null;
      minor: boolean;
      source: string;
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/getRevision

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.getRevision(1374050869);
  • getSectionOne named or numbered section of an article as plain text, without downloading the rest of the page.

    Arguments

    NameTypeRequired
    titleOrUrlstringYes
    sectionIndexstringYes
    options{ lang?: string }No

    Returns WikipediaSectionContent

    interface WikipediaSectionContent {
      title: string;
      html: string;
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/getSection

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.getSection("Espresso", "0");
  • getSectionsA Wikipedia article's sections (table of contents, outline, headings): The article's table of contents.

    Arguments

    NameTypeRequired
    titleOrUrlstringYes
    options{ lang?: string }No

    Returns { sections: WikipediaSection[]; warnings: string[] }

    interface WikipediaSection {
      index: string;
      title: string;
      level: string;  // the site's own labels, read the values off a result, never guess one from prose
      line: string;
      anchor: string;
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/getSections

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.getSections("Espresso");
  • getSummaryThe lead of an article and nothing else.

    Arguments

    NameTypeRequired
    titleOrUrlstringYes
    options{ lang?: string }No

    Returns WikipediaSummary

    interface WikipediaSummary {
      id: number;
      title: string;
      url: string;
      description: string;
      extract: string;
      extractHtml: string;
      wikidataId?: string;
      thumbnail?: {
        url: string;
        width: number;
        height: number;
      };
      originalImage?: {
        url: string;
        width: number;
        height: number;
      };
      coordinates?: {
        lat: number;
        lon: number;
      };
      revisionId: number;
      lastModified: string;
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/getSummary

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.getSummary("Espresso");
  • getUserA Wikipedia editor's public record.

    Arguments

    NameTypeRequired
    usernamesstring | string[]Yes
    options{ lang?: string }No

    Returns { users: WikipediaUser[]; warnings: string[] }

    interface WikipediaUser {
      name: string;
      editCount: number;
      registrationDate: string;
      groups: string[];
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/getUser

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.getUser("Jimbo Wales");
  • getWikitextThe article's raw wikitext source.

    Arguments

    NameTypeRequired
    titleOrUrlstringYes
    options{ lang?: string }No

    Returns WikipediaWikitext

    interface WikipediaWikitext {
      title: string;
      wikitext: string;
      revisionId: number;
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/getWikitext

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.getWikitext("Espresso");
  • listCategoriesThe Wikipedia categories an article belongs to: The categories an article belongs to.

    Arguments

    NameTypeRequired
    titleOrUrlstringYes
    options{ lang?: string; limit?: number }No

    Returns { categories: WikipediaCategory[]; warnings: string[] }

    interface WikipediaCategory {
      title: string;
      url: string;
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/listCategories

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.listCategories("Espresso");
  • listCategoryMembersEvery article in a category.

    Arguments

    NameTypeRequired
    categorystringYes
    options{ lang?: string; limit?: number }No

    Returns { members: WikipediaCategoryMember[]; warnings: string[] }

    interface WikipediaCategoryMember {
      title: string;
      url: string;
      type: "page" | "subcategory" | "file";
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/listCategoryMembers

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.listCategoryMembers("Coffee preparation");
  • listImagesEvery image, diagram, audio clip and video in an article, in page order.

    Arguments

    NameTypeRequired
    titleOrUrlstringYes
    options{ lang?: string; limit?: number }No

    Returns { images: WikipediaMediaItem[]; warnings: string[] }

    interface WikipediaMediaItem {
      title: string;
      type: string;
      section: number;
      leadImage: boolean;
      caption: string;
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/listImages

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.listImages("Espresso");
  • listLanguagesWhat this article is called in every other language Wikipedia has it in, with the url of each.

    Arguments

    NameTypeRequired
    titleOrUrlstringYes
    options{ lang?: string; limit?: number }No

    Returns { languages: WikipediaLanguageLink[]; warnings: string[] }

    interface WikipediaLanguageLink {
      lang: string;
      title: string;
      url: string;
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/listLanguages

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.listLanguages("Espresso");
  • listMostViewedWhat Wikipedia's readers looked at most.

    Arguments

    NameTypeRequired
    options{ lang?: string; limit?: number; namespace?: number; date?: string }No

    Returns { articles: WikipediaMostViewedArticle[]; warnings: string[] }

    interface WikipediaMostViewedArticle {
      title: string;
      url: string;
      views: number;
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/listMostViewed

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.listMostViewed();
  • listNotificationsThe caller's Wikipedia notifications.

    Arguments

    None.

    Returns { notifications: WikipediaNotification[]; warnings: string[] }

    interface WikipediaNotification {
      id: number;
      type: string;
      category: string;
      section: string;
      timestamp: string;
      read: boolean;
      title?: string;
      titleUrl?: string;
      agent?: string;
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/listNotifications

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.listNotifications();
  • listRecentChangesWhat is being edited on Wikipedia right now.

    Arguments

    NameTypeRequired
    options{ lang?: string; limit?: number; namespace?: number; anonymousOnly?: boolean }No

    Returns { changes: WikipediaRecentChange[]; warnings: string[] }

    interface WikipediaRecentChange {
      title: string;
      pageId: number;
      revisionId: number;
      oldRevisionId: number;
      namespace: number;
      user: string;
      anonymous: boolean;  // a classic IP editor or a temp-account placeholder, the site's own answer, never inferred from the username
      timestamp: string;
      comment: string;
      minor: boolean;
      isNew: boolean;
      bot: boolean;
      oldLength: number;
      newLength: number;
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/listRecentChanges

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.listRecentChanges({ namespace: 0, limit: 3 });
  • listRelatedArticles about things like this one.

    Arguments

    NameTypeRequired
    titleOrUrlstringYes
    options{ lang?: string; limit?: number }No

    Returns { results: WikipediaSearchResult[]; warnings: string[] }

    interface WikipediaSearchResult {
      id: number;
      title: string;
      url: string;
      excerpt: string;
      description: string;
      thumbnail?: {
        url: string;
        width: number;
        height: number;
      };
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/listRelated

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.listRelated("Espresso");
  • listRevisionsThe Wikipedia edit history (revision history, page history) of an article: An article's edit history.

    Arguments

    NameTypeRequired
    titleOrUrlstringYes
    options{ lang?: string; limit?: number }No

    Returns { revisions: WikipediaRevision[]; warnings: string[] }

    interface WikipediaRevision {
      id: number;
      timestamp: string;
      user: string;
      comment: string;
      size: number;
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/listRevisions

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.listRevisions("Espresso");
  • listUserContributionsEvery edit one named editor has made, newest first.

    Arguments

    NameTypeRequired
    usernamestringYes
    options{ lang?: string; limit?: number }No

    Returns { contributions: WikipediaUserContribution[]; warnings: string[] }

    interface WikipediaUserContribution {
      title: string;
      url: string;
      timestamp: string;
      comment: string;
      revisionId: number;
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/listUserContributions

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.listUserContributions("Jimbo Wales", { limit: 3 });
  • listWatchlistThe caller's own watchlist.

    Arguments

    None.

    Returns { items: WikipediaWatchlistItem[]; warnings: string[] }

    interface WikipediaWatchlistItem {
      title: string;
      pageId: number;
      revisionId: number;
      oldRevisionId: number;
      namespace: number;
      user: string;
      anonymous: boolean;
      timestamp: string;
      comment: string;
      minor: boolean;
      isNew: boolean;
      bot: boolean;
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/listWatchlist

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.listWatchlist();
  • searchNearbyWikipedia articles about places near a point.

    Arguments

    NameTypeRequired
    latnumberYes
    lonnumberYes
    options{ lang?: string; radius?: number; limit?: number; namespace?: number }No

    Returns { places: WikipediaNearbyPlace[]; warnings: string[] }

    interface WikipediaNearbyPlace {
      title: string;
      pageId: number;
      namespace: number;
      lat: number;
      lon: number;
      distance: number;
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/searchNearby

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.searchNearby(48.8583, 2.2944, { limit: 3 });
  • suggestTitlesAutocomplete a partial title the way Wikipedia's search box does as you type.

    Arguments

    NameTypeRequired
    querystringYes
    options{ lang?: string; limit?: number }No

    Returns { suggestions: WikipediaTitleSuggestion[]; warnings: string[] }

    interface WikipediaTitleSuggestion {
      id: number;
      title: string;
      url: string;
      description: string;
    }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/suggestTitles

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.suggestTitles("espre", { limit: 3 });
  • thankUserSend another editor Wikipedia's built-in "thanks" for a specific edit.

    Arguments

    NameTypeRequired
    revisionIdnumberYes
    options{ lang?: string }No

    Returns { thanked: true; revisionId: number; warnings: string[] }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/thankUser

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.thankUser(1);
  • undoRevisionUndo one edit to an article.

    Arguments

    NameTypeRequired
    titleOrUrlstringYes
    revisionIdnumberYes
    options{ lang?: string; summary?: string }No

    Returns { undone: true; title: string; url: string; revisionId: number; undidRevisionId: number; warnings: string[] }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/undoRevision

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.undoRevision("Espresso", 1);
  • unwatchPageRemove an article from the caller's watchlist.

    Arguments

    NameTypeRequired
    titleOrUrlstringYes
    options{ lang?: string }No

    Returns { unwatched: true; title: string; url: string; warnings: string[] }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/unwatchPage

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.unwatchPage("Espresso");
  • watchPageAdd an article to the caller's watchlist so they are told when it changes.

    Arguments

    NameTypeRequired
    titleOrUrlstringYes
    options{ lang?: string }No

    Returns { watched: true; title: string; url: string; expiry?: string; warnings: string[] }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/watchPage

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.watchPage("Espresso");
Act 3
  • createPageCreate an article that does not exist yet, under the caller's account.

    Arguments

    NameTypeRequired
    titlestringYes
    textstringYes
    options{ lang?: string; summary?: string }No

    Returns { created: true; title: string; url: string; revisionId: number; newTimestamp?: string; warnings: string[] }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/createPage

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.createPage("<title>", "<text>");
  • editPageChange an article, replace a page or one of its sections with new wikitext, under the caller's own account and with their own edit summary.

    Arguments

    NameTypeRequired
    titleOrUrlstringYes
    textstringYes
    options{ lang?: string; summary?: string; section?: string | number; minor?: boolean }No

    Returns { edited: true; title: string; url: string; revisionId: number; oldRevisionId?: number; newTimestamp?: string; warnings: string[] }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/editPage

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.editPage("<titleOrUrl>", "<text>");
  • postTalkMessagePost a message to an article's talk page or to a user's talk page, under the caller's own account.

    Arguments

    NameTypeRequired
    titleOrUrlstringYes
    messagestringYes
    options{ lang?: string; subject?: string }No

    Returns { posted: true; title: string; url: string; revisionId: number; warnings: string[] }

    HTTP endpoint

    POST https://api.bowmark.ai/v1/sites/wikipedia/postTalkMessage

    Call it

    import { bowmark } from "@bowmark/web";
    
    const result = await bowmark.providers.wikipedia.postTalkMessage("<titleOrUrl>", "<message>");

Use it from code