Example workflow — Learner

Quietly learns your preferences from agent traces with TypeSafe Jev and turns them into one-tap skill updates you approve.

Scenario

You imported Learner from the public listing and want a safe first session focused on preference learning with one-tap approved skill updates.

Steps (example)

1. Import from the verified listing and connect TypeSafe Jev only through the official flow.
2. Let the bot observe agent traces you are comfortable sharing.
3. When a skill update is proposed, read it before approving; nothing applies without your tap.
4. Revoke or adjust skills outside the bot if you change your mind later.

Approvals are intentional. Learner turns patterns into skills you explicitly accept.

Start guide · Prepare a local report