Skip to main content

What Copera AI Can Do

Ask Copera AI doesn't just answer questions — it does the work. Because your boards, documents, chat, meetings, drive, and calendar all live in one workspace, the assistant has the context to take real actions: build a tracker, create and update rows, send a message, schedule a meeting, research a topic, or assemble an automation. This page is the complete, honest reference for what it can and can't do.

Every action respects your permissions — the AI can only see and do what you can — and by default it asks before each change. See How acting works below.

Build and structure your workspace

The assistant can create the structure of your work, not just fill it in.

It can…Ask it something like
Create a board"Create a board to track our hiring pipeline."
Add a table"Add a table for candidate interviews."
Create and update columns"Add a Status column with New, Screening, Offer, Hired."
Export a table"Export the Leads table as a spreadsheet."

Work with your data

It can…Ask it something like
Create rows (one or many)"Add these five vendors as rows with their contact and renewal date."
Update rows and fields"Set every overdue task's priority to High."
Write a row's rich-text description"Write a project brief in the description of this row."
Read and search your data"Which deals are closing this month?"
Delete rows"Remove the duplicate rows in this table." (Deletes always confirm.)

Communicate and schedule

It can…Ask it something like
Send a channel message or DM"Post the release summary in the Engineering channel."
Compose or send email"Draft a follow-up email to this lead."
Schedule a meeting"Find 30 minutes with Maria this week and book it."
Check your day"What does my afternoon look like?"
Create a channel, document, or to-do"Start a channel for the launch and a doc for the plan."
Invite a member or share a document"Invite [email protected] and share the launch doc with them."

Research and create

It can…Ask it something like
Search the web"What are competitors charging for this?"
Read and analyze a web pagePaste a link and ask "summarize this page"
Run Deep Research (cited)"Do deep research on the EU AI Act's impact on SMBs."
Generate images"Create a hero image for our launch post."
Browse Copera's own help"How do formulas work in Copera?"

Automate your repetitive work

Describe an automation and the assistant plans, drafts, validates, and creates it — then you review before it runs. It can also diagnose and repair one that's failing.

"When a form is submitted, create a row in Leads and notify the assignee."

See Automations for the full engine, and the AI Playbook for 58 things you can build.

Bring your own AI tools

You can attach your own external tools (via Model Context Protocol servers) as connectors, so Ask Copera AI can call them alongside its built-in abilities. And because Copera exposes a hosted MCP server, CLI, and public API, you can also drive your Copera workspace from outside tools like Claude, Cursor, or your own scripts.

How acting works: you stay in control

The assistant's ability to act is governed by a permission mode you choose. Read and navigation actions always run; anything that changes or sends something is gated:

Permission modeWhat happens
Plan onlyThe assistant proposes what it would do but never commits a change.
Ask before acting (default)It shows a review card before every change, and you approve or reject each one.
Act except deletes (admin)It applies non-destructive changes automatically, but still confirms deletions.
Act without asking (admin)It applies changes automatically.

Two more guarantees hold in every mode:

  • It acts as you, within your permissions. The AI can only see and do what your own account can. It can't reach data you don't have access to, or take an action you're not allowed to take.
  • You choose how hard it thinks. An effort dial lets you trade speed for depth on harder, multi-step tasks.

What it can't do (yet)

Being clear about the edges matters as much as the abilities:

  • It doesn't create roles or permissions. It can navigate you to your Roles settings, but changing roles and permissions stays in your hands, on purpose.
  • It won't exceed your permissions. Ever. It's scoped to your own access.
  • Standing, autonomous teammates are early access. Today the assistant acts inside your conversation (and automations run on their triggers). Named agents that run entirely on their own schedule are rolling out gradually.
  • A few abilities are still coming, such as placing voice calls or reading a whiteboard's contents. Where a use case needs one of these, we mark it "Coming soon."
  • It doesn't train on your data. Your workspace content is used only to answer what you asked, never to train models.

Frequently Asked Questions

Can Copera AI create a board, table, or column?

Yes — it can create boards, add tables, create and update columns, and fill them with rows, so it can build a working tracker from a description. By default it shows a review card before each change.

Can Copera AI send messages and schedule meetings?

Yes — channel messages, DMs, email (compose or send), scheduling a meeting, and checking your day, plus creating channels, documents, and to-dos. It acts as you and within your permissions.

Does Copera AI act on its own without asking?

Only if you choose an act mode. The default is "Ask before acting," which shows a review card before every change. The act-automatically modes are admin-only.

Can Copera AI create roles or permissions?

No. It can take you to Roles settings, but it doesn't create or change roles itself — that stays with you.

Can Copera AI build an automation for me?

Yes — describe it and the assistant plans, drafts, validates, and creates the automation for your review, and can diagnose and repair a broken one.