AI with your rights
An AI teammate that asks before it acts.
Press ⌘J anywhere. Nova answers from your tickets, specs, time, roadmap and code, and prepares the change: a new ticket, a status, a comment, a booking. Nothing runs until you confirm, and it only sees and does what you may.
NovaET · Every Time
Worked 14 s · read 1 file
- Reading ET-11 and its pull requests
- Searching the code for “overtime export”
- Reading scan-bb/src/overtime/export.ts
Pull request #41 was opened as a draft, and drafts don’t move tickets. It left draft an hour ago, with an approving review and passing checks. exportOvertime() now writes one row per person and day, as the brief asks.
Release 2.4 · 31 open tickets
78% chance to make 27 Nov
Likely
21 Nov
Safe
28 Nov
Almost sure
3 Dec
Pace of the last 12 weeks · 7.4 tickets a weekScope 31
Gone quiet ET-14
In review for 4 days without a change
Waiting for review #57
No approving review for 3 days
Blocked ET-19
In the sprint, waits for open ET-6
Slipping 2.4
Due in 18 days, open work needs 26 at the recent pace
The assignee hears about it once; the lead when it is serious.
The app’s own components, rendered with example data.
Your session, not a superuser
Nova reads and changes the product with the session of the person asking. The database’s own row-level security, grants and triggers decide, exactly as they do in the app. When the database refuses, Nova says so and does not look for another way.
- Team-only projects stay hidden from people who are not on them, in answers too.
- Tables that keep keys and tokens are out of its reach, and so is every column that looks like a secret.
- Guests never get code.
Changes come back as a card
New tickets, edits, status, comments, time, checks and rows anywhere in the product come back as a card that shows each change before it runs: from and to, old and new. Confirm runs them as you, through the same tools agents use, so ticket permissions and workflows still apply. Discard drops them.
- Up to 12 changes per answer; untick the ones you do not want.
- Only the person who asked can confirm, and a card runs once.
It reads the code your project lives in
Nova opens the repositories assigned to a project, or linked to its tickets, and reads them at the head of their main branch. Ask why a ticket is not in review, where a rule lives, or what a change would touch, and the answer names the files.
- Reads Bitbucket repositories; GitHub and GitLab repositories are for coding agents.
- In a workspace-wide conversation it opens the code of the project a question is about, only if you can see that project.
Research, plans and the rest of the product
Auto answers quickly; Research reads widely, searches the web and cites its sources. It reaches sprints, goals, budgets, capacity, feedback, specs, automations and settings, and runs the product’s AI features as you: triage, estimates, briefs, meeting notes, standups and release notes.
- Forecasts and risks come back as cards with dates and severity.
- Planners ask for a plan and turn it into tickets from the panel.
- / opens commands, @ tags people, projects and tickets.
Questions
Is Nova an AI action?
Each answer counts as one AI action against the plan’s monthly allowance: 50 on Free, 1,000 on Pro, 5,000 on Business. With your own OpenAI key there is no limit.
Can Nova change things I cannot?
No. It uses your own session, so it can read and change exactly what you can in the app, and nothing more. Your session is used for the request only and never saved.
Are conversations private?
Yes. Conversations belong to the person who started them. A conversation started on a project can be shared with its planners from the Brainstorm tab.
Which model does it use?
OpenAI, with our key or the one you add under Integrations. Admins pick the model from a checked list, and each answer shows the model it used.
Give your next agent a brief.
Free for up to 10 people, with MCP for every agent. No credit card.