Data Studio

Data Studio is a built-in, spreadsheet-style database for each agent, a lightweight Airtable that lives inside Minicrew. You create “sheets” with columns you define, and your agent can look things up in them, or add and update rows, while it’s chatting with customers or teammates.

A product catalogue the agent can search. A bookings or waitlist sheet it adds entries to. A FAQ table it references, or a simple CRM of leads collected mid-conversation. That’s the kind of thing this is for.

Where to find it

Open an agent, go to its Knowledgebase tab, and look for the Data Sheets section. From there you can turn on “Let the assistant read & write your sheets” and create your first sheet, which opens the full Data Studio page.

Creating a sheet

1. From the Data Studio page, click + New data sheet and give it a name.
2. Add columns and choose a type for each one: text, number, currency, date, date & time, link, image, file, dropdown (select), or checkbox. Some types support a default value — for example, a date column can default to today automatically.
3. Reorder columns by dragging them if you want a particular layout.
4. Save the sheet. You can add rows one at a time, or import a batch all at once.

Importing and exporting

• Import CSV — bring in existing data (a spreadsheet export, a CSV from another tool) in bulk. Columns are matched and values converted automatically; dates, numbers, and currency get coerced into the right format.
• Export — download any sheet as a CSV or Excel (.xlsx) file at any time, from the export button on the sheet toolbar.

Controlling what the agent can do

By default an agent with Data Studio access can both read and write a sheet, but you can restrict this:

• Sheet-level permission — set a sheet to read-only for the agent, so it can look things up but never change the data.
• Column-level permission — restrict individual columns (for example, let the agent update a “status” column but never touch a “price” column).

Prompt rules

Each sheet has a Prompt rules field: plain-language instructions telling the agent when and how to use that particular sheet. “Look up the customer’s order status here before answering shipping questions,” say, or “add a new row whenever someone asks to join the waitlist.” Type a backtick (`) while writing a rule to pull up an autocomplete list of your columns.

Sheets with prompt rules set show a highlighted Prompt rules button on the toolbar, so you can tell at a glance which sheets are actively steering the agent.

Audit history

Data Studio keeps a history of changes made to a sheet, including whether the change came from the agent or from a person. Open the audit log from the sheet toolbar to check what changed and when, useful for confirming the agent is updating records correctly.

Note

Data Studio is per-agent — each agent has its own independent set of sheets. To share data across two agents, keep them in sync manually, or point both agents at the same information in their persona/rules.