fetch_link_documentation

Instructions

Implementation Reference

• src/tools/linkDocumentationTool.ts:24-41 (handler)

  The entry-point handler function for the 'fetch_link_documentation' tool. It clears visited URLs, crawls the provided URL (with optional selector and depth), formats the result, and returns it as MCP content.

  async fetchDocumentation(args: FetchDocumentationArgs) {
    this.visitedUrls.clear();
  
    try {
      const result = await this.crawlPage(args.url, args.selector || 'body', args.depth || 1);
      
      return {
        content: [
          {
            type: "text",
            text: this.formatDocumentationResult(result),
          },
        ],
      };
    } catch (error) {
      throw new Error(`Failed to fetch documentation: ${error instanceof Error ? error.message : String(error)}`);
    }
  }

• src/tools/linkDocumentationTool.ts:43-91 (helper)

  Core helper method that performs the actual web crawling, content extraction (title, text, components, APIs, examples, links), and recursive crawling based on depth.

  private async crawlPage(url: string, selector: string, depth: number): Promise<DocumentationResult> {
    if (this.visitedUrls.has(url) || depth < 0) {
      return this.createEmptyResult(url);
    }
  
    this.visitedUrls.add(url);
  
    try {
      const response = await axios.get(url, {
        timeout: 10000,
        headers: {
          'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36'
        }
      });
  
      const $ = cheerio.load(response.data);
      const selectedContent = $(selector);
  
      const result: DocumentationResult = {
        url,
        title: $('title').text().trim() || $('h1').first().text().trim() || 'Untitled',
        content: this.extractTextContent(selectedContent),
        components: this.extractComponents($),
        apis: this.extractAPIs($),
        examples: this.extractExamples($),
        links: this.extractLinks($, url),
      };
  
      // 如果depth > 0，继续爬取相关链接
      if (depth > 0) {
        const relatedLinks = result.links
          .filter(link => this.isRelevantLink(link))
          .slice(0, 5); // 限制每页最多5个链接
  
        for (const link of relatedLinks) {
          const subResult = await this.crawlPage(link, selector, depth - 1);
          result.content += `\n\n--- Linked Page: ${subResult.title} ---\n${subResult.content}`;
          result.components.push(...subResult.components);
          result.apis.push(...subResult.apis);
          result.examples.push(...subResult.examples);
        }
      }
  
      return result;
    } catch (error) {
      console.error(`Error crawling ${url}:`, error);
      return this.createEmptyResult(url);
    }
  }

• src/index.ts:39-58 (schema)

  JSON schema defining the input parameters for the tool: required 'url', optional 'selector' and 'depth'.

  inputSchema: {
    type: "object",
    properties: {
      url: {
        type: "string",
        description: "The URL to fetch documentation from (e.g., WindUI documentation)",
      },
      selector: {
        type: "string",
        description: "Optional CSS selector to focus on specific content",
        default: "body"
      },
      depth: {
        type: "number",
        description: "How deep to crawl links (0-3, default: 1)",
        default: 1
      }
    },
    required: ["url"],
  },

• src/index.ts:33-62 (registration)

  Registers the tool in MCP ListToolsRequestHandler, providing name, description, and input schema.

  this.server.setRequestHandler(ListToolsRequestSchema, async () => {
    return {
      tools: [
        {
          name: "fetch_link_documentation",
          description: "Fetch and analyze documentation from a website link, extracting all available components, APIs, and usage examples",
          inputSchema: {
            type: "object",
            properties: {
              url: {
                type: "string",
                description: "The URL to fetch documentation from (e.g., WindUI documentation)",
              },
              selector: {
                type: "string",
                description: "Optional CSS selector to focus on specific content",
                default: "body"
              },
              depth: {
                type: "number",
                description: "How deep to crawl links (0-3, default: 1)",
                default: 1
              }
            },
            required: ["url"],
          },
        },
      ],
    };
  });

• src/index.ts:68-71 (registration)

  Dispatches calls to 'fetch_link_documentation' to the LinkDocumentationTool instance in the CallToolRequestHandler.

  switch (name) {
    case "fetch_link_documentation":
      return await this.linkTool.fetchDocumentation(args as any);