Dumpling AI MCP Server

Instructions

Implementation Reference

• src/index.ts:701-722 (handler)

  The handler function that implements the tool logic by proxying requests to the external Dumpling AI API endpoint for extracting structured data from images.

  async ({ inputMethod, images, prompt, jsonMode }) => {
    const apiKey = process.env.DUMPLING_API_KEY;
    if (!apiKey) throw new Error("DUMPLING_API_KEY not set");
    const response = await fetch(`${NWS_API_BASE}/api/v1/extract-image`, {
      method: "POST",
      headers: {
        "Content-Type": "application/json",
        Authorization: `Bearer ${apiKey}`,
      },
      body: JSON.stringify({
        inputMethod,
        image: images[0],
        prompt,
        jsonMode,
        requestSource: "mcp",
      }), // Assuming single image for simplicity
    });
    if (!response.ok)
      throw new Error(`Failed: ${response.status} ${await response.text()}`);
    const data = await response.json();
    return { content: [{ type: "text", text: JSON.stringify(data, null, 2) }] };
  }

• src/index.ts:693-700 (schema)

  Zod input schema defining parameters for the extract-image tool: inputMethod, images array, prompt, and optional jsonMode.

  {
    inputMethod: z.enum(["url", "base64"]).describe("Input method"),
    images: z
      .array(z.string())
      .describe("Array of URLs or base64-encoded images"),
    prompt: z.string().describe("Extraction prompt"),
    jsonMode: z.boolean().optional().describe("Return in JSON format"),
  },

• src/index.ts:690-723 (registration)

  Registration of the 'extract-image' MCP tool using server.tool, including name, description, input schema, and handler.

  server.tool(
    "extract-image",
    "Extract structured data from images based on a prompt.",
    {
      inputMethod: z.enum(["url", "base64"]).describe("Input method"),
      images: z
        .array(z.string())
        .describe("Array of URLs or base64-encoded images"),
      prompt: z.string().describe("Extraction prompt"),
      jsonMode: z.boolean().optional().describe("Return in JSON format"),
    },
    async ({ inputMethod, images, prompt, jsonMode }) => {
      const apiKey = process.env.DUMPLING_API_KEY;
      if (!apiKey) throw new Error("DUMPLING_API_KEY not set");
      const response = await fetch(`${NWS_API_BASE}/api/v1/extract-image`, {
        method: "POST",
        headers: {
          "Content-Type": "application/json",
          Authorization: `Bearer ${apiKey}`,
        },
        body: JSON.stringify({
          inputMethod,
          image: images[0],
          prompt,
          jsonMode,
          requestSource: "mcp",
        }), // Assuming single image for simplicity
      });
      if (!response.ok)
        throw new Error(`Failed: ${response.status} ${await response.text()}`);
      const data = await response.json();
      return { content: [{ type: "text", text: JSON.stringify(data, null, 2) }] };
    }
  );