generated: '2026-08-01'
method: searched
source: https://support.datarails.com/hc/en-us
description: >-
  Datarails' client-side surface is unusual: instead of embeddable web
  components for third-party developers, it ships host-application add-ins —
  Excel and PowerPoint — plus agent-side connectors that embed Datarails data
  into AI clients. Both are genuine client-side components, but neither is a
  developer-embeddable element library, and no loader script or component CDN
  is published.

families:
  - family: Office add-ins
    description: First-party add-ins that render live Datarails data inside Microsoft Office.
    components:
      - name: Datarails Flex (Excel add-in)
        surface: Microsoft Excel
        platforms: [Windows, macOS]
        docs:
          - https://support.datarails.com/hc/en-us/articles/15716070391964-Datarails-Flex-for-Windows-Installation
          - https://support.datarails.com/hc/en-us/articles/15716095917596-Datarails-Flex-for-Mac-Installation
        capabilities:
          - DR.GET formulas that pull live financial data into cells
          - Drill-down from a formula cell to underlying line items
          - Filters and named ranges
          - Planning mode
        note: >-
          DR.GET is Datarails' in-cell query language and is effectively a
          client-side data-binding syntax. The plugin repository documents a
          DR.GET authoring contract for agents that generate workbooks.
      - name: PowerPoint add-in
        surface: Microsoft PowerPoint
        docs: https://support.datarails.com/hc/en-us/articles/4418562347793-PowerPoint-Add-in-Installation
        capabilities:
          - Embed Datarails-sourced charts and figures into decks
      - name: Connect Office JS filters
        surface: Microsoft Excel (Office.js)
        docs: https://support.datarails.com/hc/en-us/articles/26458280440732-Connect-Office-JS-Filters
        note: >-
          The add-in exposes an Office.js execution path; the first-party agent
          skills reference an execute_office_js tool for driving it.

  - family: AI client connectors
    description: >-
      Packaged bindings that surface Datarails data inside third-party AI
      clients. Client-side in the sense that they run in the user's assistant,
      not in a Datarails-hosted page.
    components:
      - name: Datarails FinanceOS Claude Code plugin
        client: Claude Code / Claude Desktop / Claude CoWork
        version: 3.0.6
        url: https://github.com/Datarails/dr-claude-code-plugins-re
        see: skills/_index.yml
      - name: Datarails FinanceOS ChatGPT app
        client: ChatGPT (Plus, Team, Enterprise)
        docs: https://support.datarails.com/hc/en-us/articles/25905041804444-Connect-Datarails-to-ChatGPT
        note: Listed in the ChatGPT App Directory; connects over MCP with OAuth.
      - name: Datarails FP&A Copilot Agent for Microsoft 365
        client: Microsoft 365 Copilot
        docs: https://support.datarails.com/hc/en-us/articles/26842914887452-Set-Up-Your-Datarails-FP-A-Copilot-Agent-in-Microsoft-365
      - name: Datarails on Claude for Excel
        client: Claude for Excel
        docs: https://support.datarails.com/hc/en-us/articles/25874100800668-Getting-Started-with-Datarails-on-Claude-for-Excel

  - family: In-product surfaces
    description: >-
      Datarails-hosted UI that customers configure rather than embed. Recorded
      for completeness; these are not developer components.
    components:
      - name: Dashboards
        docs: https://support.datarails.com/hc/en-us/articles/19714914749084-Dashboards-Overview
      - name: Report Builder
        docs: https://support.datarails.com/hc/en-us/articles/18981849298972-Report-Builder-Overview
      - name: Web Editor for Collaborators
        docs: https://support.datarails.com/hc/en-us/articles/29082848881820-Web-Editor-For-Collaborators

low_code:
  - name: Lovable
    docs: https://support.datarails.com/hc/en-us/articles/26384776911644-Create-Datarails-Related-Apps-with-Lovable
    note: >-
      Datarails documents building apps against FinanceOS through Lovable, and
      a lovable_verification TXT record exists on datarails.com. This is the
      closest Datarails comes to a third-party app-building path.

gaps:
  - No embeddable web components, custom elements or JS element library for third-party sites.
  - No published component CDN or loader script.
  - No design-system package intended for external consumption (the @datarailsshared npm packages are first-party internals).