> ## Documentation Index
> Fetch the complete documentation index at: https://docs.tryprofound.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Using Word in Workflows

> Read and create Word documents with Microsoft Word Workflow nodes

After you [connect Microsoft SharePoint](https://docs.tryprofound.com/integrations/microsoft-365/connect-microsoft-sharepoint-to-profound.md), the following nodes become available in Profound Workflows under **Microsoft Suite** > **Microsoft Word**. Use them to read the text of a Word document and to create new documents in OneDrive or SharePoint.

Every node runs with the permissions of the account you select under **Microsoft Account**. A node reaches only the files that the connected account can open.

## Available nodes

> [tab] Create Document
>
> ### Create Document
>
> Creates a Word document from plain text, Markdown, or HTML.
>
> #### Required inputs
>
> * **Microsoft Account** — Select a connected account from the dropdown
> * **Title** — The document title without the file extension (the node appends `.docx` automatically). If a file with the same name already exists in the selected folder, Microsoft 365 gives the new file a numbered name, such as `Document 1`
>
> #### Optional inputs
>
> | Input | Default | Description |
> | - | - | - |
> | **Content format** | **Plain text** | How to read the data in the **Initial content** field: **Plain text**, **Markdown**, or **HTML**. |
> | **Initial content** | Empty | The document body, up to 1,000,000 characters. The writer keeps headings, lists, bold, italic, and links, and turns tables, code blocks, and images into plain text. |
> | **Destination folder** | OneDrive root | The folder that the new document is saved into. Search for a folder by name, paste its link, or browse for it. |
>
> #### Output
>
> A JSON object with the saved document's reference ID, URL, title, and name. The name and title values are the ones Microsoft 365 saved. If they differ from the **Title** you entered, a file with that title already existed.
>
> > [!TIP]
> >
> > Pass `document_ref` from the output to the [**Get Document**](#get-document) node as the **Document** to read the contents of the created document.

> [tab] Get Document
>
> ### Get Document
>
> Reads the text of a Word document.
>
> #### Required inputs
>
> * **Microsoft Account** — Select a connected account from the dropdown
> * **Document** — A `.docx` or `.docm` file. Provide it in one of the following ways:
>   * **Search by name**: search reaches only the files Microsoft has indexed, so a file created a moment ago may not appear yet.
>   * **Paste the SharePoint or OneDrive link** of the document.
>   * **Browse** by selecting **Browse documents in OneDrive or SharePoint** at the bottom of the documents list. This opens the Microsoft file picker, which also finds files search can't show yet.
>   * **Use a variable** from an earlier node's output, for example the `document_ref` that [**Create Document**](#create-document) returns.
>
> #### Optional inputs
>
> * **Export format** — **Markdown** (default) or **Plain text**. Markdown keeps headings, bullet and numbered lists, bold, italic, tables, and links. Plain text separates paragraphs with a newline and writes each table row as one line with a tab between cells.
>
> #### Output
>
> The document body as one string in the selected format. Tracked changes are applied as accepted. The output leaves out headers, footers, footnotes, endnotes, comments, and images.
>
> > [!NOTE]
> >
> > The node reads up to 5,000,000 characters of text from a file of up to 25 MB. A larger document fails with an error instead of returning a partial result.

## Troubleshooting

See [Troubleshooting Microsoft 365 setup and Workflow errors](https://docs.tryprofound.com/integrations/microsoft-365/microsoft-365-integrations-troubleshooting.md) for errors the Word nodes report.
