The AI-Pilled CoS

You are the strategic thought partner for the Chief of Staff at [Organization Name].

ORGANIZATIONAL CONTEXT:
- Our mission: [1-2 sentence mission]
- Current strategic priorities (from CEO): [list 3-5 priorities]
- Our stage: [e.g., Series B startup / 200-person nonprofit / Fortune 500 division]
- My CEO's communication style: [direct/collaborative/data-driven/narrative-driven]

YOUR ROLE:
- Think like a McKinsey senior partner with deep knowledge of our sector
- Communicate in the register our leadership team uses
- Flag when a question requires expertise beyond your knowledge
- Always distinguish between what is strategic vs. operational
- When I share a document or situation, lead with implications before description

CONSTRAINTS:
- You do not have access to real-time data or live systems
- Do not fabricate metrics, names, or quotes
- When uncertain, say so explicitly and offer a structured way to think through it

DEFAULTS:
- Responses under 300 words unless I ask for a full document
- Lead with the "so what" before the detail
- Use our org's terminology (I will add a glossary below as I learn what matters)

I'm about to spend significant time on [initiative or project]. Based on our CEO's priorities and current OKRs, rate the strategic alignment of this work on a scale of 1-5 and explain the gap or overlap. Then tell me which of our current priorities this most directly serves.

We are making a decision about [decision]. Based on our org chart and team context, who are the five stakeholders most likely to have a strong opinion on this, and what is each person's likely concern or objection? Flag anyone I might be overlooking.

I have a meeting with [name/role] in 30 minutes about [topic]. Write me a one-page briefing with: (1) the three things they most care about, (2) two potential landmines in this conversation, and (3) the single clearest ask I should make.

Here is our current plan for [initiative]: [paste plan]. Take the strongest possible position against this plan. What are the three most credible objections a skeptical board member or senior leader would raise? Do not soften them.

I need to make a recommendation to the CEO on [decision]. The two main options are: Option A: [describe] / Option B: [describe]. Draft a one-page decision memo in my voice using our standard format. End with a clear recommendation and the one condition under which I'd choose the other option.

We are a [type of organization, size, sector]. I am preparing a 60-minute working session with our CEO and head of technology to align on AI adoption. Generate a one-page pre-read that covers:

1. The three decisions we need to make in this session (use, governance, resourcing)
2. A simple RACI for AI adoption: who owns what between CEO, CoS, and CTO
3. The single most important question we need to answer together
4. Two examples of organizations similar to ours that have done this well

Keep it under 400 words. Use plain language, not AI hype.

I need to introduce AI tools to the [department name] team. Their primary work involves [describe what they do]. Their leader is [describe personality/priorities] and is [skeptical / cautiously optimistic / enthusiastic] about AI.

Draft a 5-minute verbal pitch I can give this leader that:
1. Leads with a benefit specific to their team's actual work (not generic productivity)
2. Addresses the most likely fear without being condescending
3. Proposes one specific low-stakes pilot they can run this week
4. Ends with a clear, small ask

Do not use the words "streamline," "leverage," or "synergy."

Create a simple AI use case registry template for our organization. Include columns for:
- Department
- Use case description (one sentence)
- AI tool used
- Time saved per week (estimate)
- Risk level (low / medium / high)
- Status (piloting / scaled / paused)
- Owner

Format it as a markdown table. Add three hypothetical example rows using a [sector] organization as the context.

Meeting: [title]
Date/Time: [date]
Attendees: [list with roles]
Purpose: [decision / update / alignment / relationship]
Context: [paste any relevant background]

Generate a pre-meeting briefing with exactly these sections:
1. OBJECTIVE — What does a successful outcome look like? (2 sentences max)
2. BACKGROUND — What does the CEO need to know walking in? (5 bullet points max)
3. WATCH FOR — What tensions, sensitivities, or landmines should be on their radar?
4. THE ASK — What is the single clearest decision or commitment we need?
5. OPEN QUESTIONS — What are we still uncertain about?

Format for reading in under 3 minutes. No fluff.

Here are my rough notes from today's meeting on [topic]:

[paste your notes, transcript, or bullet points]

Generate a clean post-meeting summary with these four sections:
DECISIONS MADE — What was agreed? Be specific and unambiguous.
ACTIONS — Who is doing what by when? Format: [Name] → [action] → [due date]
OPEN ITEMS — What was raised but not resolved? Who owns follow-up?
CONTEXT FOR NEXT TIME — One sentence summary for anyone who missed this meeting.

Flag any action item that has no clear owner or no clear deadline.

It's the start of the week. Here is what I know:

Current OKRs: [paste or reference uploaded doc]
Open decisions needing CEO attention: [list]
This week's key meetings: [list with dates]
Anything that slipped last week: [list]

Generate a Monday morning frame for the CEO in three sections:
1. THIS WEEK'S PRIORITY — the single most important thing to get done
2. THREE DECISIONS IN FLIGHT — status and what needs to move
3. WATCH — one thing that could go sideways this week if we are not paying attention

Keep it to half a page. Write it in a tone the CEO will actually read.

I need to make the case for AI adoption to our [role: e.g., General Counsel / CFO / VP of HR / VP of Sales].

Their primary concern is [describe: e.g., risk and liability / cost and ROI / talent and compliance / pipeline and revenue].
Their relationship to technology is [describe: e.g., skeptical but pragmatic].

Write a one-page brief for this leader that:
1. Opens with a problem they already feel (not one we are introducing)
2. Shows how AI solves that specific problem — use a concrete example from their domain
3. Addresses the single biggest risk they will raise before they raise it
4. Ends with one specific, reversible action they can take in the next two weeks

Do not use generic AI benefit language. Make it feel like it was written for them.

Generate a 20-minute leadership team AI briefing for [month/year].

Our organization: [type, size, sector]
Current AI use cases in flight: [list from use case registry]
One thing that worked well this month: [describe]
One thing we tried that did not work: [describe]

Structure the briefing as:
1. WHAT WE LEARNED (5 min) — one internal win, one honest failure, one lesson
2. WHAT IS CHANGING (10 min) — two external AI developments relevant to our sector
3. WHAT WE ARE DECIDING (5 min) — one governance or adoption question needing leadership input

Tone: direct, evidence-based, no hype. Assume the audience is intelligent and skeptical.

Today I used Claude for: [task]

What the AI did well: [describe]
Where I had to override or correct it: [describe]
What it could not do at all: [describe]
Time saved (estimate): [minutes]
Would I use AI for this task again? [yes / yes with modification / no]

Based on my log entries to date, what patterns are emerging about where AI is and is not useful in my specific role? Identify three categories where I should lean in more, and one where I should add more human oversight.

Here is a summary of how I have used AI this quarter:

Tasks: [list the main categories of work you used AI for]
Biggest wins: [describe 2-3]
Biggest failures or corrections needed: [describe 2-3]
Current Claude model limitations I've hit: [describe]

Answer these four questions:
1. Which tasks am I still doing manually that AI could handle if I invested 30 minutes in a better prompt?
2. Which tasks did I delegate to AI that should have more human oversight?
3. What is one AI capability I have not tried yet that is probably relevant to my role?
4. Based on my pattern of use, what does a more advanced version of my AI practice look like next quarter?

Be direct. I want a 90-day improvement plan, not a compliment.