Dumpling AI MCP Server

Instructions

Implementation Reference

• src/index.ts:1029-1051 (handler)

  The handler function that implements the logic for the 'run-python-code' tool by proxying the request to the Dumpling AI API endpoint for executing Python code in a sandbox.

  async ({ code, commands, parseJson, timeout, memory, saveOutputFiles }) => {
    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/run-python-code`, {
      method: "POST",
      headers: {
        "Content-Type": "application/json",
        Authorization: `Bearer ${apiKey}`,
      },
      body: JSON.stringify({
        code,
        commands,
        parseJson,
        timeout,
        memory,
        saveOutputFiles,
      }),
    });
    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:1022-1028 (schema)

  Zod schema defining the input parameters for the 'run-python-code' tool including code, optional pip commands, JSON parsing flag, timeout, memory, and file saving options.

    code: z.string().describe("Python code"),
    commands: z.string().optional().describe("Pip install commands"),
    parseJson: z.boolean().optional().describe("Parse output as JSON"),
    timeout: z.number().optional().describe("Execution timeout in ms"),
    memory: z.number().optional().describe("Memory allocation in MB"),
    saveOutputFiles: z.boolean().optional().describe("Save output files"),
  },

• src/index.ts:1018-1021 (registration)

  Registration of the 'run-python-code' tool using server.tool() with its name, description, schema, and handler function.

  server.tool(
    "run-python-code",
    "Execute Python code in a sandbox.",
    {