Dataset Viewer MCP Server

Instructions

Implementation Reference

• src/dataset_viewer/server.py:547-557 (handler)

  MCP tool handler in @server.call_tool() that extracts arguments (dataset, config, split), instantiates DatasetViewerAPI with auth_token, calls its get_statistics method, formats the result as JSON text content, and returns it.

  elif name == "get_statistics":
      dataset = arguments["dataset"]
      config = arguments["config"]
      split = arguments["split"]
      stats = await DatasetViewerAPI(auth_token=auth_token).get_statistics(dataset, config=config, split=split)
      return [
          types.TextContent(
              type="text",
              text=json.dumps(stats, indent=2)
          )
      ]

• src/dataset_viewer/server.py:82-91 (helper)

  Core implementation in DatasetViewerAPI class that constructs parameters and makes asynchronous HTTP GET request to the Hugging Face dataset viewer /statistics endpoint, returning the JSON response.

  async def get_statistics(self, dataset: str, config: str, split: str) -> dict:
      """Get statistics about a dataset"""
      params = {
          "dataset": dataset,
          "config": config,
          "split": split
      }
      response = await self.client.get("/statistics", params=params)
      response.raise_for_status()
      return response.json()

• src/dataset_viewer/server.py:388-414 (schema)

  Input schema definition for the get_statistics tool, specifying required parameters (dataset, config, split) with types, descriptions, patterns, examples, and optional auth_token.

  inputSchema={
      "type": "object",
      "properties": {
          "dataset": {
              "type": "string",
              "description": "Hugging Face dataset identifier in the format owner/dataset",
              "pattern": "^[^/]+/[^/]+$",
              "examples": ["ylecun/mnist", "stanfordnlp/imdb"]
          },
          "config": {
              "type": "string",
              "description": "Dataset configuration/subset name. Use get_info to list available configs",
              "examples": ["default", "en", "es"]
          },
          "split": {
              "type": "string",
              "description": "Dataset split name. Splits partition the data for training/evaluation",
              "examples": ["train", "validation", "test"]
          },
          "auth_token": {
              "type": "string",
              "description": "Hugging Face auth token for private/gated datasets",
              "optional": True
          }
      },
      "required": ["dataset", "config", "split"],
  }

• src/dataset_viewer/server.py:385-415 (registration)

  Registration of the get_statistics tool in the @server.list_tools() handler, including name, description, and full input schema.

  types.Tool(
      name="get_statistics",
      description="Get statistics about a Hugging Face dataset",
      inputSchema={
          "type": "object",
          "properties": {
              "dataset": {
                  "type": "string",
                  "description": "Hugging Face dataset identifier in the format owner/dataset",
                  "pattern": "^[^/]+/[^/]+$",
                  "examples": ["ylecun/mnist", "stanfordnlp/imdb"]
              },
              "config": {
                  "type": "string",
                  "description": "Dataset configuration/subset name. Use get_info to list available configs",
                  "examples": ["default", "en", "es"]
              },
              "split": {
                  "type": "string",
                  "description": "Dataset split name. Splits partition the data for training/evaluation",
                  "examples": ["train", "validation", "test"]
              },
              "auth_token": {
                  "type": "string",
                  "description": "Hugging Face auth token for private/gated datasets",
                  "optional": True
              }
          },
          "required": ["dataset", "config", "split"],
      }
  ),