A MCP server to connect to wolfram alpha API.

Components

Prompts

This is analogous to the !wa bang in duckduckgo search.

def wa(query: str) -> f"Use wolfram alpha to answer the following question: {query}"

Tools

Query Wolfram Alpha api.

def query_wolfram_alpha(query: str) -> str

Configuration

You must set the WOLFRAM_API_KEY environment variable. Get an api ket from Wolfram Alpha.

This was tested with the full results API, but it might not be required.

{
    "mcpServers": {
        "MCP-wolfram-alpha": {
            "command": "uv",
            "args": [
                "--directory",
                "C:UsersrootDocumentsMCP-wolfram-alpha",
                "run",
                "MCP-wolfram-alpha"
            ],
            "env": {
                "WOLFRAM_API_KEY": "your-app-id"
            }
        }
    }
}

Development

Debugging

Since the official MCP inspector does not have good environment support, I reccommend using wong2's mcp-cli-inspector.

Create a config.json file in the same style as claude desktop.

{
    "mcpServers": {
        "MCP-wolfram-alpha": {
            "command": "uv",
            "args": [
                "--directory",
                "/full/path/to/MCP-wolfram-alpha",
                "run",
                "MCP-wolfram-alpha"
            ],
            "env": {
                "WOLFRAM_API_KEY": "your-app-id"
            }
        }
    }
}

Then run:

npx @wong2/mcp-cli -c .config.json
[
  {
    "description": "Use Wolfram Alpha to answer a question. This tool should be used when you need complex math or symbolic intelligence.",
    "inputSchema": {
      "properties": {
        "query": {
          "type": "string"
        }
      },
      "required": [
        "query"
      ],
      "type": "object"
    },
    "name": "query-wolfram-alpha"
  }
]