Your first bot, in plain English.

No tech background needed. One small win at a time. Explore first. Learn first. It’s OK if it breaks.

Four steps to your first helper

  1. Pick one thing. Bills, email or a health habit. Just one.
  2. Give read-only access to just that one thing. It can look. It can't change anything.
  3. Tell it what you want in your own words. Like you'd ask a friend.
  4. Look at what it finds before you act. You make the call.
Screenshot of real Unbundle report files from a test run on a practice folder: a numbered preview list of twelve dummy files, largest first, and three scratch files listed for you to handle.
What “look before you act” looks like: a preview list, nothing changed, waiting on your yes. Real files from a test run on a practice folder.
Try Unbundle, the free skill

If you get stuck

You will. I did. Ask a friend. Ask me. Hold the rope for the next person when it’s your turn. That’s the whole idea: we lift others while we climb.

About your data, honestly

What I share: my bills, so a bot can compare them to what my card offers. The money connection is read-only. It can't change billing.

What I don't: my logins. The bots can't sign in and make a switch for me. Switching a card, changing a plan or signing up for a benefit needs my yes, and I push the button myself.

Turning it off: access is narrow on purpose, one thing at a time, so I can take it back the same way.

It isn't risk-free, and I won't tell you it is. This is just what I actually do.

Plain questions

Is it safe to let AI look at my bills?

It isn't risk-free, and I won't tell you it is. I keep it narrow: read-only access to one thing, no logins, and I make every change myself.

What should my first AI helper do?

Pick one thing that bugs you. Bills, email or a health habit. Just one.

Can I take access back later?

Yes. Access is narrow on purpose, one thing at a time, so I can take it back the same way.