ANIMA Workspace · Auto Mode

Auto Mode starts the moment a task arrives and gets better with every run.

Mentions, assignments, email and schedules set it off. Agents do the work and leave the key calls to you.

Auto Mode is on: Claude Code takes a Jira ticket and stops at review; a Slack question is answered; a customer quote is ready; the weekly report is ready before the sync.

Wherever work shows up

A ticket is assigned to you: Claude Code starts within the limits you set and opens a PR for your review. A teammate mentions you: your ANIMA answers within the scope you set. A customer emails for a quote: agents prepare the quote and reply from your pricing rules and similar past deals. Before the weekly meeting: agents gather updates from every source and prepare the report. Regular check-ins: agents check for stuck or overdue work and follow up with the owner.

Once it's on, it runs 24/7.

Teammates get answers without waiting on you.

Your ANIMA answers teammates within the scope you've opened up. Topics you've already approved go out on their own. Anything else waits for you.

Teammates' questions sort themselves: the ship date and the discount rule are approved topics, so your ANIMA answers them; a question about cutting a contract price waits for you.

It runs on its own until it needs you.

Jira ticket PROJ-418 comes in. Claude Code brings in the #eng thread, the last similar fix (PR #402) and the coding guidelines, fixes the export function, passes all 12 tests, opens PR #518 and stops for your review.

It learns how you work and does it for you next time.

Company context: the returns policy, with sources. Learns from corrections: you change a damaged-item reply to ship a replacement first, and replies to 3 more cases are drafted the same way for your OK. Written back to the knowledge base: the returns SOP gains the new rule and moves from v5 to v6, and every agent uses the update.

Company context

Agents work from your company context: the rules, standards and past decisions in your knowledge base.

Learns from corrections

What you approve, change or skip becomes the basis for next time. Correct it once and it does it your way.

Updates the knowledge base

What it learns is written back into your knowledge base, so every agent works from the update.

Routine work done. Key calls left to you.

Low risk (internal and easy to undo, like research, notes and status updates): agents just do it and keep a log. Medium risk (affects your team, like replying to teammates, updating tickets or opening a PR): done within the scope you've approved, then you're told. High risk (leaves the company or is hard to undo, like quotes, outside emails and contracts): prepared, then waits for your approval.

Every step is logged: the sources used, what was done and who approved it.

Save your spot

Join the waitlist. ANIMA Workspace launches on Mac first.