Give it your context. Get an ally.

People call this "cloning yourself." That's the wrong way to think about it. You're not making a copy of yourself. You're giving your assistant enough real context that it stops answering like a stranger.

Without this, even a well-briefed assistant answers every question from zero. It doesn't know your priorities this quarter, the decisions you keep making the same way, or the handful of documents you open every day.

The shortcut: let it write the first draft

You've probably already explained a lot of this in past chats. Use that. Open your assistant and paste this in:

Look back through our previous conversations and tell me what you've learned about how I work, what I've built, and what I keep coming back to. Write it up as a short document covering: my role and priorities, my track record, the documents and links I rely on most, and the patterns in how I make decisions.

Correct it, then save it

The first draft won't be perfect. Fix what's wrong, the same way you corrected your voice guide earlier in this series. Save the final version as a file in your project, or paste it straight into your instructions.

Keep one rule: one owner (you), and update it whenever something real changes. Five minutes, not a project.

Where this goes

Do this, and the assistant from this whole series stops feeling like a tool and starts feeling like an ally. It's also exactly what I build for entire companies, just pointed at one person instead of a whole team. If that's useful for your business, book a free 30-minute call. No pitch, just a clear conversation.