get_document_details

Instructions

Implementation Reference

• src/index.ts:351-395 (registration)

  Registration of the 'get_document_details' MCP tool. Defines the tool name, description, input schema (document 'nummer'), and handler function that fetches HTML from the document page and extracts structured details using extractDocumentDetailsFromHtml.

    "get_document_details",
    "Retrieves metadata about a parliamentary document in a structured JSON format, without downloading the actual document content. Returns information including title, type, document number, dates, version number, and links to both the PDF version and the official Tweede Kamer webpage.",
    {
      nummer: z.string().describe("Document number (e.g., '2024D39058') - the unique identifier for the parliamentary document you want information about")
    },
    async ({ nummer }) => {
      try {
        const html = await apiService.fetchHtml(`/document.html?nummer=${encodeURIComponent(nummer)}`);
        if (!html) {
          return {
            content: [{
              type: "text",
              text: JSON.stringify({ error: `No content found for document ${nummer}` })
            }]
          };
        }
  
        const details = extractDocumentDetailsFromHtml(html, BASE_URL);
        if (!details) {
          return {
            content: [{
              type: "text",
              text: JSON.stringify({ error: `Failed to parse details for document ${nummer}` })
            }]
          };
        }
  
        return {
          content: [{
            type: "text",
            text: JSON.stringify(details, null, 2)
          }]
        };
      } catch (error: any) {
        return {
          content: [{
            type: "text",
            text: JSON.stringify({
              error: `Error fetching document details: ${error.message || 'Unknown error'}`
            })
          }]
        };
      }
    }
  );

• src/index.ts:356-394 (handler)

  Handler function for get_document_details tool: fetches the document.html page using apiService, extracts details using the helper function, formats as JSON, and returns as MCP content or error response.

  async ({ nummer }) => {
    try {
      const html = await apiService.fetchHtml(`/document.html?nummer=${encodeURIComponent(nummer)}`);
      if (!html) {
        return {
          content: [{
            type: "text",
            text: JSON.stringify({ error: `No content found for document ${nummer}` })
          }]
        };
      }
  
      const details = extractDocumentDetailsFromHtml(html, BASE_URL);
      if (!details) {
        return {
          content: [{
            type: "text",
            text: JSON.stringify({ error: `Failed to parse details for document ${nummer}` })
          }]
        };
      }
  
      return {
        content: [{
          type: "text",
          text: JSON.stringify(details, null, 2)
        }]
      };
    } catch (error: any) {
      return {
        content: [{
          type: "text",
          text: JSON.stringify({
            error: `Error fetching document details: ${error.message || 'Unknown error'}`
          })
        }]
      };
    }
  }

• src/index.ts:353-355 (schema)

  Input schema using Zod: requires 'nummer' string parameter for the document ID.

  {
    nummer: z.string().describe("Document number (e.g., '2024D39058') - the unique identifier for the parliamentary document you want information about")
  },

• src/utils/html-parser.ts:119-178 (helper)

  Core helper function that implements the document details extraction logic by parsing HTML with regex patterns to pull title, type, nummer, datum, bijgewerkt, versie, direct PDF link, Tweede Kamer link, and attachment info (bijlageBij). Uses shared extractValue utility.

  export function extractDocumentDetailsFromHtml(html: string, baseUrl: string): DocumentDetails | null {
    if (!html) {
      return null;
    }
  
    const details: DocumentDetails = {
      title: null,
      type: null,
      nummer: null,
      datum: null,
      bijgewerkt: null,
      versie: null,
      directLinkPdf: null,
      tweedekamerLink: null,
      bijlageBij: null,
    };
  
    // Extract basic info
    details.title = extractValue(html, /<hblock>\s*<h2>([\s\S]*?)<\/h2>/i);
    details.type = extractValue(html, /<\/hblock>[\s\S]*?<p><em>([\s\S]*?)<\/em><\/p>/i);
  
    // Extract metadata line
    const metadataMatch = html.match(/<p>Nummer: <b>(.*?)<\/b>, datum: <b>(.*?)<\/b>, bijgewerkt: <b>(.*?)<\/b>, versie: (\d+)/i);
    if (metadataMatch) {
      details.nummer = metadataMatch[1]?.trim() || null;
      details.datum = metadataMatch[2]?.trim() || null;
      details.bijgewerkt = metadataMatch[3]?.trim() || null;
      details.versie = metadataMatch[4] ? parseInt(metadataMatch[4], 10) : null;
    }
  
    // Extract links
    const directLinkMatch = html.match(/<a href="(getraw\/[^"']+)">Directe link naar document<\/a>/i);
    if (directLinkMatch && directLinkMatch[1]) {
      // Ensure the URL includes the /tkconv/ path
      const rawPath = directLinkMatch[1];
      // If baseUrl already includes /tkconv/, this will work correctly
      // If not, we need to add it manually
      if (baseUrl.endsWith('/tkconv')) {
        details.directLinkPdf = `${baseUrl}/${rawPath}`;
      } else if (baseUrl.includes('/tkconv/')) {
        details.directLinkPdf = new URL(rawPath, baseUrl).href;
      } else {
        // Ensure we have /tkconv/ in the path
        details.directLinkPdf = `${baseUrl}/tkconv/${rawPath}`;
      }
    }
    details.tweedekamerLink = extractValue(html, /<a href="(https:\/\/www\.tweedekamer\.nl\/[^"']+)">link naar pagina op de Tweede Kamer site<\/a>/i);
  
    // Extract bijlage bij info
    const bijlageMatch = html.match(/<p>Bijlage bij: <a href="(document\.html\?nummer=[^"]+)">([\s\S]*?)<\/a> \(([^)]+)\)<\/p>/i);
    if (bijlageMatch && bijlageMatch[1]) {
      details.bijlageBij = {
        title: bijlageMatch[2]?.trim() || null,
        nummer: bijlageMatch[3]?.trim() || null,
        link: new URL(bijlageMatch[1], baseUrl).href,
      };
    }
  
    return details;
  }

• src/utils/html-parser.ts:5-19 (schema)

  TypeScript interface defining the structure of the document details object returned by the tool's extraction logic.

  interface DocumentDetails {
    title: string | null;
    type: string | null;
    nummer: string | null;
    datum: string | null;
    bijgewerkt: string | null;
    versie: number | null;
    directLinkPdf: string | null;
    tweedekamerLink: string | null;
    bijlageBij: {
      title: string | null;
      nummer: string | null;
      link: string;
    } | null;
  }