YOUR AGENT
NEEDS MORE
THAN APROMPT.

I want you to interview me so you can understand who I am and how to work with me. Ask me questions one at a time, waiting for my answer before continuing. Cover: - My professional role and what I'm currently building - My biggest priorities and challenges right now - How I like to communicate and receive information - My working style and how I make decisions - What success looks like for me in the next 90 days After the interview, write a profile summary and confirm it with me. Then use everything you've learned to personalize every response going forward. Start with your first question.

Based on everything you know about me so far, tell me what you're most likely to get wrong. Where are you guessing about me rather than actually knowing? What have you assumed that I never told you? Then ask me about the gaps that matter most, one at a time.

You already know who I am. Now interview me about the work itself rather than about me. Ask one question at a time. Cover what this work actually is, who it's for, what success looks like, what constrains it, and what makes my approach to it different. When you're done, write it up as context I can drop into a Project so I never have to explain it from scratch again.

I'm going to describe how I spend a typical week. For each thing I name, tell me whether it's work you could take over completely, work you could speed up but where I still need to make the call, or work that should stay entirely mine. Ask me questions where you need more detail before deciding. At the end, tell me which one you'd start with and why.

Here's how I sorted my work into things I'd hand over, things I'd speed up, and things I'd keep entirely. Argue with me. Tell me which ones I've put in the wrong bucket and why, especially anything I claimed only I can do. Be direct about it.

Here's an idea I've been sitting on. Tell me honestly whether it's buildable, what it would actually take, and what the simplest version that still delivers most of the value would look like. If it isn't realistic, say so plainly rather than being encouraging.

Here's a job I need done. Tell me whether this is something we should work through together in a conversation, or something I should hand off to you to complete on your own. Explain which parts of it need me in the room and which don't.

Based on what you know about me and my work, suggest the Projects I should set up. For each one, tell me what belongs in its instructions, what files I should put in it, and what kinds of tasks would live there. Keep it to the handful I'd actually use.

Interview me about this body of work so you can write the instructions for its Project. Ask one question at a time. I want them short enough that I'd actually read them: what this work is, who it's for, and what matters about it.

Here's what my work looks like day to day and here are the tools I use. Tell me which ones to connect first, and for each one, what specifically you'd be able to do for me once it's connected. Rank them by how much difference they'd make.

Now that you're connected, tell me three things you can see about my work that I probably haven't noticed myself. Then pick the most useful one and show me.

I'm on the entry-level paid plan and I want to be smart about it. Here's the work I do in a typical week. Tell me which of these jobs actually need your deepest thinking and which would be fine with a lighter approach, and how I should sequence my week so I'm not burning the heavy stuff on things that don't need it.

I want to automate this. Walk me through setting it up one step at a time, and check with me before each step. Before we schedule anything, tell me what could go wrong and what you'd do if you weren't sure about something.

Read my website as a first-time visitor who has never heard of me. Where do you get confused? What would stop you filling out that form? Read this proposal as the person who has to approve the spend. What's your first objection? Read this listing as a buyer. What's missing that you'd have to call and ask about?

Go through the last thirty posts on my Instagram profile. Tell me which hooks actually stopped the scroll, what my content mix looks like by format, how consistent my posting is, what's repeating too often, and whether my call to action is doing anything. Then tell me the three things you'd change.

Look at my website and the materials I've given you. Pull out my colors, my fonts, the feel of the brand in one sentence, and most importantly the things I should never do. Ask me about anything you're guessing at rather than assuming.

Based on everything we did today and what you now know about my work, write me a 90 day plan: the first 30 days, 31 to 60, and 61 to 90. Be specific, and keep it to things I'd realistically do rather than things that sound good. Then be honest with me about three things: which parts of this I'll realistically do myself, which parts I'll start and abandon, and which parts genuinely need someone else's hands. Don't be encouraging about it.

I want to write down how I sound so I stop having to explain it every time. Here are three or four things I've written: [paste sent emails, Slack messages, or meeting notes]. Work out what's actually distinctive about how I write. Sentence length, how I open, how I close, how formal I am, what I never do, how I handle bad news. Tell me what you notice before you write anything. Then write it up as a skill I can reuse. Include when it should fire, and be a little pushy about that. And add a line at the end that makes you check your own work before handing it to me.

I want a skill that stress-tests my thinking rather than agreeing with me. Interview me first. Ask one question at a time. Find out what I tend to overlook, where I've been wrong before, what kind of feedback actually lands with me and what I dismiss, and what my blind spots are when I'm excited about something. Then write it as a skill that takes whatever I bring it and tells me what I haven't considered, what I'm assuming without evidence, and where this falls apart. It should push. Being encouraging is not the job. Fire it whenever I'm working through a decision or a plan, even if I don't ask for it.

I want a skill that looks at what's come in and tells me what actually needs me. Interview me first. Ask one question at a time. Find out what makes something urgent for me versus just loud, what I always deal with personally, what I ignore for a day without consequence, who gets a fast response no matter what, and what a good morning looks like when I've handled the right things. Then write it as a skill that goes through my inbox and messages and gives me a short ranked list: what needs me now, what needs me today, what can wait, and what I should just delete. Tell me why for each one. Keep the list short. A long list is the problem I'm trying to solve.

I want a skill that reviews my work the way I would if I were being tough about it. Interview me about what I check before [a proposal / a client email / a deliverable] goes out. Ask one question at a time. Cover what I look for, what I've been burned by before, what I'd hand straight back if a junior person gave it to me, and what I let slide because it doesn't matter. Then write it as a skill that takes a draft and tells me what's wrong with it, ranked by how much it actually matters. Be harder on me than you want to be. A review that says it's fine is a review that wasted my time.

Here are notes from two meetings I've run, and the follow-up emails I sent afterward: [paste them]. Work out what I actually pull out of a meeting and how I structure a follow-up. What I always include, what I leave out, how direct I am about deadlines. Then write it as a skill that takes raw meeting notes and gives me back the decisions, the open items with who owns them and by when, and a drafted follow-up email. Ask me anything you need before you write it.

I want to write down how I decide whether something is worth my time. Interview me about how I qualify [a lead / an inquiry / a new opportunity]. Ask one question at a time. Get at what I check, the order I check it in, what makes me walk away immediately, and the exceptions I make that contradict my own rules. Push me when my answer is vague. Then write it as a skill that takes a new inquiry and gives me a recommendation with the reasoning. Tell me which parts of my logic you're least confident you got right.

Here are the last two versions of a report I produce every [week / month]: [paste or attach them]. Work out its structure, what I lead with, what I always include, and the level of detail I use. Then write it as a skill that takes my raw notes and data and produces the next one in the same shape. Tell me exactly what inputs you need from me each time.

Here are two proposals I've sent, ideally one that closed and one that didn't: [paste or attach them]. Work out my structure, how I present scope, how I handle pricing and terms, and what I always include. Then tell me what was different about the one that didn't close. Then write it as a skill that takes rough scope notes from a call and produces a proposal in my format. Flag anything you're guessing at rather than seeing in both examples.

I want a skill that reads my work as the person on the other side of it, not as me. Interview me about who those people usually are. Ask one question at a time. Find out who reads my proposals, who lands on my website, who I'm usually trying to convince, what they already know, what they're skeptical about, and what would make them walk away. Then write it as a skill where I tell it which of those people to be, and it reads whatever I give it as that person. It should tell me where it gets confused, what it doesn't believe, and what would stop it acting. Do not soften it.

I want my visual standards written down as a skill so they apply everywhere, not just in Design. Here's what I've got: [paste your brand kit, or a link to your website, or attach your brand guide]. Pull out my colors, my type, the feel in one sentence, and most importantly the things I should never do. Ask me about anything you're guessing at rather than assuming. Then write it as a skill that applies these whenever I'm making something that gets seen: a document, a one-pager, a deck, a page. Fire it without being asked.