get_page

Instructions

Implementation Reference

• src/index.js:264-297 (handler)

  The main handler for the 'get_page' tool. Accepts optional 'url' or 'path', fetches the page HTML via fetchPage(), extracts readable content, and returns title, URL, path, content_text, content_markdown, and links.

  server.tool(
    'get_page',
    'Obtiene una p�gina concreta del sitio.',
    {
      url: z.string().optional(),
      path: z.string().optional()
    },
    async ({ url, path }) => {
      try {
        if (!url && !path) throw new Error('INVALID_URL');
        const target = url || path;
        const { html, finalUrl } = await fetchPage(target);
        const extracted = extractReadableContent(html, finalUrl);
  
        return {
          content: [
            {
              type: 'text',
              text: JSON.stringify({
                title: extracted.title,
                url: finalUrl,
                path: new URL(finalUrl).pathname,
                content_text: extracted.contentText,
                content_markdown: `# ${extracted.title}\n\n${extracted.contentText}`,
                links: extracted.links
              }, null, 2)
            }
          ]
        };
      } catch (error) {
        return mcpError(error);
      }
    }
  );

• src/index.js:267-270 (schema)

  Input schema for 'get_page': accepts optional 'url' (string) and 'path' (string) parameters.

  {
    url: z.string().optional(),
    path: z.string().optional()
  },

• src/index.js:264-265 (registration)

  Registration of 'get_page' tool via server.tool(), with description 'Obtiene una página concreta del sitio.'.

  server.tool(
    'get_page',

• src/index.js:198-207 (helper)

  Helper function fetchPage() that ensures authentication, makes a GET request, and returns the HTML and final URL.

  async function fetchPage(urlOrPath) {
    await ensureAuthenticated();
    const response = await request(urlOrPath, { method: 'GET' });
    if (!response.ok) {
      throw new Error(response.status === 404 ? 'PAGE_NOT_FOUND' : 'INTERNAL_ERROR');
    }
    const finalUrl = response.url || normalizeUrl(urlOrPath);
    const html = await response.text();
    return { html, finalUrl };
  }

• src/index.js:114-164 (helper)

  Helper function extractReadableContent() that parses HTML with cheerio, extracts title, sections, links, and cleaned body text.

  function extractReadableContent(html, pageUrl) {
    const $ = cheerio.load(html);
    $('script, style, noscript').remove();
  
    const title = $('title').first().text().trim() || $('h1').first().text().trim() || pageUrl;
    const sections = [];
  
    $('h1, h2, h3').each((_, el) => {
      const heading = $(el).text().trim();
      const texts = [];
      let current = $(el).next();
  
      while (current.length && !['h1', 'h2', 'h3'].includes(current.get(0)?.tagName)) {
        const text = current.text().trim();
        if (text) texts.push(text);
        current = current.next();
      }
  
      if (heading || texts.length) {
        sections.push({ heading, text: texts.join('\n\n') });
      }
    });
  
    const links = [];
    $('a[href]').each((_, el) => {
      const href = $(el).attr('href');
      const label = $(el).text().trim();
      if (!href) return;
      try {
        const url = new URL(href, pageUrl).toString();
        if (url.startsWith(BASE_URL)) {
          links.push({ label: label || url, url });
        }
      } catch {
      }
    });
  
    const contentText = $('body')
      .text()
      .replace(/\s+\n/g, '\n')
      .replace(/\n\s+/g, '\n')
      .replace(/\n{3,}/g, '\n\n')
      .trim();
  
    return {
      title,
      contentText,
      sections,
      links: dedupeLinks(links)
    };
  }