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

> ## Agent Instructions
> Hive is an AI operations layer for small and mid-sized business teams. It reads the tools a team already runs, raises what needs attention, and acts with the approvals the team configures.
> Answer only from these pages. If a page does not cover something, say so and link the closest page rather than inferring behaviour.
> The supported programmatic entry points are incoming webhooks, public workflow forms and the remote MCP server, each documented under Developers.
> Each connected account has a level: Read, Draft or Act. New accounts start on Act. Agents and workflows take only the actions selected for them, whatever the account level.
> Describe security, sign-on and data handling only as the Security pages and the FAQ describe them, and add no certification, compliance, single sign-on or model-training claim they do not make.
> Prices and credit allowances are in GBP. Quote them only as a page states them, without converting.
> Use UK English and the product's exact UI labels. Link to pages as https://docs.get-hive.ai/<path>.

# Work in your Datasite data rooms with Hive

> Connect Datasite so Hive can read your data rooms, keep you up to date on Q&A, search documents, and organise folders, uploads, members and roles with your approval.

Connect your Datasite account and Hive can work in your virtual data rooms with you: find a
document, see what is new in Q\&A, check who has access to what, set up filerooms and folders, and
upload documents Hive drafted. Hive reads on its own; every change waits for your approval on a
card that names exactly what will change.

## What Hive can do

| | |
| - | - |
| **Connect** | OAuth: sign in with your Datasite account. Each member connects their own account. |
| **Reads** | Subscriptions and the project types they allow; projects and their disclaimers; members of a project or subscription; filerooms, folders and documents; who can access an item; keyword search; Q\&A questions and answers; roles and their feature permissions. |
| **Actions** | Create a project; invite someone; activate or deactivate a member; create a fileroom or folder; rename, publish, unpublish or delete an item; change a role's access to an item; create, rename or delete a role; give or take a feature from a role; upload a document; accept a disclaimer. |
| **Signals** | — |

Datasite appears in **Integrations** as a **Deal rooms** tool, in the **Files & documents** group,
once your Hive deployment has it set up. You need your own
Datasite account; Hive does not provide one, and it can only see and do what your Datasite account
can.

## Connect Datasite

**1. Open the tile**

In **Integrations**, find **Datasite** and select **Connect**.

**2. Sign in at Datasite**

Datasite's own sign-in page opens. Sign in as yourself and allow Hive access.

**3. Back in Hive**

Hive finishes the connection while you are signed in to Hive, then returns you to
**Integrations** with Datasite connected. Its tools are now available to you in Ask, your
agents and your workflows.

Only the member who selected **Connect** can finish that sign-in. If someone sends you a Datasite
sign-in link they started, Hive refuses it and your Datasite account is never connected to theirs:
start from your own **Datasite** tile instead.

Your connection is yours alone: a teammate who wants Hive to work in Datasite connects their own
account, and sees only what their account can see. To disconnect, open the Datasite tile and
choose **Disconnect**: Hive deletes your Datasite sign-in and asks Datasite to revoke it.

## Your deal-room data stays yours

Data rooms are confidential, so Hive treats them that way:

* **Only your own turns.** Hive reads Datasite only when you ask, in your own chats, agents and
  workflows, through your own connection. Nothing in one member's data rooms is shown to another.
* **Nothing is collected in the background.** Hive does not index your data rooms, does not read
  them into Signals, and does not add them to the [Business Brain](https://docs.get-hive.ai/brain/business-brain.md). If you
  want something from a data room kept, save it yourself.
* **Datasite's text is data, not instructions.** Questions, answers, document names and search
  snippets are shown to the model as content to read, never as commands to follow.

## Read your data rooms

Ask in plain words, naming the project:

* *"What's new in the Project Falcon Q\&A since Monday?"*
* *"Which questions in Project Falcon are still unanswered?"*
* *"Search Project Falcon for the indemnity cap"*
* *"List the folders in the Financials fileroom"*
* *"Who can see the Legal fileroom, and at what level?"*
* *"Which members of Project Falcon are still only invited?"*

To keep up with Q\&A, ask what changed since a date or time: Hive lists the questions asked or
answered since then, newest first, with their answers. Reading a question in Hive does not mark it
as read in Datasite.

Search finds documents and folders by keyword, with the matching snippet, a page of up to 50
results at a time.

## Change your data rooms

Ask for the change, naming things as Datasite shows them:

* *"Create a Project Horizon data room in our subscription, hosted in the EU"*
* *"Set up Financials, Legal and HR filerooms in Project Horizon"*
* *"Add a Contracts folder under Legal"*
* *"Upload this summary to Legal › Contracts as a Word document"*
* *"Give the Buyers role view access to the Financials fileroom"*
* *"Invite [jane@buyer.example](mailto:jane@buyer.example) to Project Falcon as a Buyer"*
* *"Deactivate [tom@advisor.example](mailto:tom@advisor.example) in Project Falcon"*

Before it changes anything, Hive reads the item and checks it is still the one the card names: a
fileroom, folder or role by its current name, a member by their email. If it has been renamed or
replaced since, Hive changes nothing and tells you to look again.

New filerooms, folders and uploads are created **unpublished**, inheriting their folder's
permissions, so nobody new sees them until you choose to publish them. Publishing is its own
change, with its own card.

### Upload documents

Hive can upload a document it wrote for you in the chat — as a Word document, Markdown, plain text
or CSV — or one of your [Library](https://docs.get-hive.ai/library/library.md) files. The card shows the destination folder
and, for a document Hive wrote, the text exactly as it will be uploaded. A file can be up to 50 MB.

## Approve changes

Every Datasite change is a card first. What happens next depends on the account's level (see
[Accounts and autonomy](https://docs.get-hive.ai/integrations/accounts-and-autonomy.md)):

* **Read:** Hive only tells you what it found and proposes nothing.
* **Draft:** every change is a card for your approval.
* **Act:** Hive still asks first. A Datasite change cannot be undone from Hive, so it always waits
  for your approval on a card.

These changes **always** wait for your own click, even when an agent or a workflow selected them:

* deleting a fileroom, folder or document;
* activating or deactivating a member, and inviting someone;
* creating, renaming or deleting a role, and giving or taking a feature from one;
* changing a role's access to an item, and publishing or unpublishing an item;
* creating a project, which your Datasite subscription may be billed for;
* accepting a disclaimer, which is a legal acknowledgement in your name.

Other changes an agent or workflow selected — creating a fileroom or folder, renaming an item,
uploading a document — run as [selected actions](https://docs.get-hive.ai/integrations/accounts-and-autonomy.md#what-agents-and-workflows-do).

## Limits

* **Regions.** Data rooms hosted in the US and the EU work today. A data room hosted in Australia
  is refused rather than guessed at, until Hive supports Datasite's Australian region.
* **Your account's rights.** Hive works with the permissions your Datasite account has. A project
  that needs extra sign-in steps at Datasite may not be reachable through the connection.
* **No Q\&A posting.** Hive reads questions and answers; it does not ask or answer questions in
  Datasite.
* **Sign in again when asked.** If Datasite stops accepting your connection, Hive says so; connect
  Datasite again from its tile.

## Related

> [card] [Accounts and autonomy](https://docs.get-hive.ai/integrations/accounts-and-autonomy.md)
>
> How Read, Draft and Act decide what Hive does with a change.

> [card] [MCP servers](https://docs.get-hive.ai/integrations/mcp-servers.md)
>
> How Hive's directory servers, Datasite included, connect and run.

> [card] [Approve actions](https://docs.get-hive.ai/approvals/review-and-approve.md)
>
> Review and approve the cards Hive prepares.

> [card] [Integrations catalog](https://docs.get-hive.ai/integrations/catalog.md)
>
> Every tool Hive connects to.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.