Wolfram|Alpha · Capability

Wolfram|Alpha Computational Search

Comprehensive computational search capability combining the Full Results API (complete pod-based results) and LLM API (AI-optimized results) for applications requiring deep computational intelligence. Used by search platforms, educational tools, and research applications needing rich multi-format results.

Run with Naftiko ComputationEducationFull ResultsMathResearchScienceSearchWolfram Alpha

What You Can Do

GET
Compute full — Get complete Wolfram|Alpha results with all pods and formats.
/v1/compute
GET
Compute for llm — Get Wolfram|Alpha results formatted for LLM processing.
/v1/compute/llm

MCP Tools

compute-full-results

Get complete Wolfram|Alpha computational results with all data pods. Returns structured JSON with math, science, geography, and more.

read-only
compute-for-llm

Query Wolfram|Alpha and receive computational results formatted for LLM consumption. Includes structured text with interpretation and result data.

read-only

APIs Used

wolfram-full wolfram-llm

Capability Spec

computational-search.yaml Raw ↑
naftiko: "1.0.0-alpha1"

info:
  label: "Wolfram|Alpha Computational Search"
  description: >-
    Comprehensive computational search capability combining the Full Results API
    (complete pod-based results) and LLM API (AI-optimized results) for applications
    requiring deep computational intelligence. Used by search platforms, educational
    tools, and research applications needing rich multi-format results.
  tags:
    - Computation
    - Education
    - Full Results
    - Math
    - Research
    - Science
    - Search
    - Wolfram Alpha
  created: "2026-05-03"
  modified: "2026-05-03"

binds:
  - namespace: env
    keys:
      WOLFRAM_ALPHA_APP_ID: WOLFRAM_ALPHA_APP_ID

capability:
  consumes:
    - import: wolfram-full
      location: ./shared/full-results-api.yaml
    - import: wolfram-llm
      location: ./shared/llm-api.yaml

  exposes:
    - type: rest
      port: 8081
      namespace: wolfram-search-api
      description: "Unified REST API for Wolfram|Alpha computational search."
      resources:
        - path: /v1/compute
          name: computations
          description: "Full computational results with pod-based output."
          operations:
            - method: GET
              name: compute-full
              description: "Get complete Wolfram|Alpha results with all pods and formats."
              call: "wolfram-full.query-full-results"
              with:
                input: "rest.input"
                output: "rest.output"
                format: "rest.format"
                includepodid: "rest.includepodid"
                units: "rest.units"
              outputParameters:
                - type: object
                  mapping: "$."

        - path: /v1/compute/llm
          name: llm-computations
          description: "LLM-optimized computational results."
          operations:
            - method: GET
              name: compute-for-llm
              description: "Get Wolfram|Alpha results formatted for LLM processing."
              call: "wolfram-llm.query-llm-api"
              with:
                input: "rest.input"
                maxchars: "rest.maxchars"
              outputParameters:
                - type: object
                  mapping: "$."

    - type: mcp
      port: 9091
      namespace: wolfram-search-mcp
      transport: http
      description: "MCP server for AI-assisted computational search and scientific queries."
      tools:
        - name: compute-full-results
          description: >-
            Get complete Wolfram|Alpha computational results with all data pods.
            Returns structured JSON with math, science, geography, and more.
          hints:
            readOnly: true
            openWorld: true
          call: "wolfram-full.query-full-results"
          with:
            input: "tools.input"
            output: "tools.output"
            format: "tools.format"
          outputParameters:
            - type: object
              mapping: "$."

        - name: compute-for-llm
          description: >-
            Query Wolfram|Alpha and receive computational results formatted
            for LLM consumption. Includes structured text with interpretation
            and result data.
          hints:
            readOnly: true
            openWorld: true
          call: "wolfram-llm.query-llm-api"
          with:
            input: "tools.input"
            maxchars: "tools.maxchars"
          outputParameters:
            - type: object
              mapping: "$."