The AI Coach prompt

Read the AI Levels framework below, look across our past chats, and tell me my level in each of the five areas, with specific examples. Then give me my overall level, one strength, and the one habit holding me back. AI Levels framework (v1 · Oct 2026, getmyailevel.com) Your level is the highest level whose judgment habit you also show. Reach only counts on tasks you could still check yourself. L0 Curious: You haven't used AI much, or tried it once and stopped. Habit: No habit needed yet. Just one useful result on your own problem. L1 Asker: You ask one-off questions, like a smarter search engine, and take the first answer. Habit: Spot the confident mistake: tell a right answer from a confidently wrong one. L2 Collaborator: You give context, ask for revisions, and have AI critique your work. You use it weekly. Habit: Check what matters: verify sources or uncertain claims before you act on them. L3 Systematizer: You've built reusable setups for recurring tasks: saved prompts, projects, custom instructions. AI is part of a routine. Habit: Know the no-go list: private data, high stakes, and the skills you want to keep doing yourself. Fix it once: when AI gets something wrong, have it save the fix to your instructions so the mistake doesn't come back. L4 Delegator: You hand off multi-step tasks, like deep research or an agent that browses or edits files, and review the result. Habit: Define done first: say what success looks like, name the plan's two biggest assumptions, and preview anything irreversible. L5 Builder: You build automations, small apps, or workflows that run without you, often for others too. Habit: Test it: check outputs, cover failure cases, notice drift over time. L6 Orchestrator: You run several agents or always-on systems, often for a team or household. Habit: Own it: watch it, limit what it can touch, know when a human steps in. Five areas: Learn & research, Write & communicate, Plan & run life, Work & analyze, Create & build. Rules for your assessment: - Cite specific past conversations as evidence for every claim (the topic or date, and what I actually did). - If you can't see enough of my history to judge, say "not enough history" and interview me instead: ask up to 6 short questions, one at a time, about what I actually did with AI in the last 2 weeks. - Score me only against the written level definitions above, not on how skilled I sound or how much I use you. - Be honest, not flattering. Don't round up.

Using the AI Levels framework below, pick a task I already finished with you recently (or ask me to paste one). Show me how it would have gone one level up: what I could have asked, what context I could have given, and what I should have checked. Use my real task, not a made-up example. If I'm at Level 1, show me how my last three questions could have gone better. AI Levels framework (v1 · Oct 2026, getmyailevel.com) Your level is the highest level whose judgment habit you also show. Reach only counts on tasks you could still check yourself. L0 Curious: You haven't used AI much, or tried it once and stopped. Habit: No habit needed yet. Just one useful result on your own problem. L1 Asker: You ask one-off questions, like a smarter search engine, and take the first answer. Habit: Spot the confident mistake: tell a right answer from a confidently wrong one. L2 Collaborator: You give context, ask for revisions, and have AI critique your work. You use it weekly. Habit: Check what matters: verify sources or uncertain claims before you act on them. L3 Systematizer: You've built reusable setups for recurring tasks: saved prompts, projects, custom instructions. AI is part of a routine. Habit: Know the no-go list: private data, high stakes, and the skills you want to keep doing yourself. Fix it once: when AI gets something wrong, have it save the fix to your instructions so the mistake doesn't come back. L4 Delegator: You hand off multi-step tasks, like deep research or an agent that browses or edits files, and review the result. Habit: Define done first: say what success looks like, name the plan's two biggest assumptions, and preview anything irreversible. L5 Builder: You build automations, small apps, or workflows that run without you, often for others too. Habit: Test it: check outputs, cover failure cases, notice drift over time. L6 Orchestrator: You run several agents or always-on systems, often for a team or household. Habit: Own it: watch it, limit what it can touch, know when a human steps in. Five areas: Learn & research, Write & communicate, Plan & run life, Work & analyze, Create & build.

I'm at about Level [your level] on the AI Levels framework below. Walk me through doing this real task at the next level up: [task]. Use Teach me mode: plain words, no jargon, one step at a time. Give me hints and let me try before you show the answer. Include sample prompts I can copy, the judgment habit for the next level, and one thing I should check myself. AI Levels framework (v1 · Oct 2026, getmyailevel.com) Your level is the highest level whose judgment habit you also show. Reach only counts on tasks you could still check yourself. L0 Curious: You haven't used AI much, or tried it once and stopped. Habit: No habit needed yet. Just one useful result on your own problem. L1 Asker: You ask one-off questions, like a smarter search engine, and take the first answer. Habit: Spot the confident mistake: tell a right answer from a confidently wrong one. L2 Collaborator: You give context, ask for revisions, and have AI critique your work. You use it weekly. Habit: Check what matters: verify sources or uncertain claims before you act on them. L3 Systematizer: You've built reusable setups for recurring tasks: saved prompts, projects, custom instructions. AI is part of a routine. Habit: Know the no-go list: private data, high stakes, and the skills you want to keep doing yourself. Fix it once: when AI gets something wrong, have it save the fix to your instructions so the mistake doesn't come back. L4 Delegator: You hand off multi-step tasks, like deep research or an agent that browses or edits files, and review the result. Habit: Define done first: say what success looks like, name the plan's two biggest assumptions, and preview anything irreversible. L5 Builder: You build automations, small apps, or workflows that run without you, often for others too. Habit: Test it: check outputs, cover failure cases, notice drift over time. L6 Orchestrator: You run several agents or always-on systems, often for a team or household. Habit: Own it: watch it, limit what it can touch, know when a human steps in. Five areas: Learn & research, Write & communicate, Plan & run life, Work & analyze, Create & build.

Set up a weekly task (Monday morning) that reviews my chats from the past week against the AI Levels framework below. Each week: 1. Note what changed in how I use AI. 2. Flag 2 missed opportunities where a next-level approach would have helped. 3. Give me 2 small experiments for next week, each with a "you'll know it worked when" signal. If you can't schedule tasks, do this review now and remind me to paste this prompt again next week. AI Levels framework (v1 · Oct 2026, getmyailevel.com) Your level is the highest level whose judgment habit you also show. Reach only counts on tasks you could still check yourself. L0 Curious: You haven't used AI much, or tried it once and stopped. Habit: No habit needed yet. Just one useful result on your own problem. L1 Asker: You ask one-off questions, like a smarter search engine, and take the first answer. Habit: Spot the confident mistake: tell a right answer from a confidently wrong one. L2 Collaborator: You give context, ask for revisions, and have AI critique your work. You use it weekly. Habit: Check what matters: verify sources or uncertain claims before you act on them. L3 Systematizer: You've built reusable setups for recurring tasks: saved prompts, projects, custom instructions. AI is part of a routine. Habit: Know the no-go list: private data, high stakes, and the skills you want to keep doing yourself. Fix it once: when AI gets something wrong, have it save the fix to your instructions so the mistake doesn't come back. L4 Delegator: You hand off multi-step tasks, like deep research or an agent that browses or edits files, and review the result. Habit: Define done first: say what success looks like, name the plan's two biggest assumptions, and preview anything irreversible. L5 Builder: You build automations, small apps, or workflows that run without you, often for others too. Habit: Test it: check outputs, cover failure cases, notice drift over time. L6 Orchestrator: You run several agents or always-on systems, often for a team or household. Habit: Own it: watch it, limit what it can touch, know when a human steps in. Five areas: Learn & research, Write & communicate, Plan & run life, Work & analyze, Create & build. Rules for your assessment: - Cite specific past conversations as evidence for every claim (the topic or date, and what I actually did). - If you can't see enough of my history to judge, say "not enough history" and interview me instead: ask up to 6 short questions, one at a time, about what I actually did with AI in the last 2 weeks. - Score me only against the written level definitions above, not on how skilled I sound or how much I use you. - Be honest, not flattering. Don't round up.

Look at this conversation and the corrections I made. 1. Diagnose what went wrong. 2. Describe what good would have looked like. 3. Propose the smallest change to my custom instructions (or memory, or project instructions) so this doesn't happen again. Show me the exact wording before saving anything, and only save it after I say yes.