Development Tools MCP Server

Instructions

Implementation Reference

• src/tools/web-scraping.ts:302-309 (handler)

  Main handler dispatch for 'extract_links' tool within handleWebScrapingTool function. Chooses between dynamic and static scraper based on useBrowser flag and returns the links.

  case 'extract_links': {
    if (config.useBrowser) {
      const data = await dynamicScraper.scrapeDynamicContent(config);
      return data.links;
    } else {
      return await staticScraper.extractLinks(config);
    }
  }

• src/tools/web-scraping.ts:56-72 (schema)

  Input schema definition for the 'extract_links' tool.

  name: 'extract_links',
  description: 'Extract all links from a web page',
  inputSchema: {
    type: 'object',
    properties: {
      url: {
        type: 'string',
        description: 'URL to scrape',
      },
      useBrowser: {
        type: 'boolean',
        description: 'Use browser for dynamic content',
        default: false,
      },
    },
    required: ['url'],
  },

• src/tools/web-scraping.ts:55-73 (registration)

  Registration of the 'extract_links' tool in the webScrapingTools array, which is later included in the MCP server's allTools.

  {
    name: 'extract_links',
    description: 'Extract all links from a web page',
    inputSchema: {
      type: 'object',
      properties: {
        url: {
          type: 'string',
          description: 'URL to scrape',
        },
        useBrowser: {
          type: 'boolean',
          description: 'Use browser for dynamic content',
          default: false,
        },
      },
      required: ['url'],
    },
  },

• src/scrapers/static-scraper.ts:126-129 (handler)

  Static scraper implementation of extractLinks: invokes scrapeHTML to parse page and returns the links array.

  async extractLinks(config: ScrapingConfig): Promise<string[]> {
    const data = await this.scrapeHTML(config);
    return data.links || [];
  }

• src/scrapers/static-scraper.ts:32-43 (helper)

  Core logic for extracting links using Cheerio: iterates over <a> tags, resolves relative URLs, and collects unique hrefs.

  const links: string[] = [];
  $('a[href]').each((_, element) => {
    const href = $(element).attr('href');
    if (href) {
      try {
        const url = new URL(href, config.url);
        links.push(url.href);
      } catch {
        // Invalid URL, skip
      }
    }
  });

• src/server.ts:18-25 (registration)

  Top-level registration: includes webScrapingTools (containing extract_links) into the allTools list provided to the MCP server for tool listing.

  const allTools = [
    ...codeAnalysisTools,
    ...codeQualityTools,
    ...dependencyAnalysisTools,
    ...lintingTools,
    ...webScrapingTools,
    ...apiDiscoveryTools,
  ];