Copilot
Command Center

First prompts and basic navigation for new Copilot users. This guide covers requirements, first wins, the prompt formula, and where to find Copilot across Microsoft 365.

Requirements

• Microsoft 365 account (free Copilot Chat available at copilot.microsoft.com)
• Copilot Premium or Copilot for Microsoft 365 for inbox, calendar, and document access
• Access via web, desktop apps, or mobile

Where to Find Copilot

Your First 5 Wins

1. Summarize a Long Email Thread

Select the thread in Outlook → click Copilot → paste:

Summarize this thread in 3 bullet points: what's the issue, what's been decided, what needs my action.

2. Draft a Reply

Select an email → click Copilot → paste:

Draft a reply that acknowledges the request, provides a clear next step, and matches the sender's tone.

3. Summarize a Document

Open any Word doc or PDF → click Copilot → paste:

Give me a 5-bullet summary: main purpose, key findings, recommendations, risks, and what needs my attention.

4. Create a Meeting Agenda

In Copilot Chat → paste:

Create a 60-minute meeting agenda for [topic]. Include time blocks, discussion items, decision points, and assign ownership for each item.

5. Rewrite for Clarity

Select any text → click Copilot → paste:

Rewrite this to be clearer and more concise. Use active voice, shorter sentences, and put the main point first.

The Prompt Formula

Every effective Copilot prompt has three parts:

[ROLE] + [TASK] + [FORMAT]

Examples:

Full prompt example:

Act as my executive assistant. Review my unread emails from today. Categorize them into action required, FYI, and noise. Present as a table with columns: Priority, From, Subject, Category, Suggested Action.

Prompting Best Practices

1. Be specific: "Summarize in 5 bullets" beats "Summarize this"
2. Set the role: "Act as a [role]" calibrates tone and depth
3. Define the format: Tables, bullets, paragraphs — specify your preference
4. Provide context: Attach files, select text, or paste background info
5. Iterate: Follow up with "Make it shorter", "Add more detail on X", or "Rewrite in a different tone"
6. Use follow-ups: Copilot remembers context within a chat session — build on previous outputs

Common Mistakes to Avoid

• Vague prompts: "Help me with this email" → too broad, produces generic output
• No format specification: Copilot defaults to paragraphs; request tables or bullets if you prefer them
• Not selecting context: In Outlook/Word, highlight the relevant content before clicking Copilot
• One-and-done: The first output is rarely perfect — iterate with follow-up prompts
• Ignoring web search: For current information, add "Search the web for the latest on..." to your prompt

Next Steps

1. Set up your profile: Configure Copilot memory with your role and preferences (see Copilot Personalization)
2. Try the starter prompts: Run through the 10 Starter Prompts to build familiarity
3. Establish your daily rhythm: Use Start My Day each morning and Weekly Case Study Review each Friday
4. Explore advanced prompts: Browse the full Prompt Library for role-specific and situational prompts

The first 10 prompts every new Copilot user should try. These cover the most common daily tasks — from summarizing your day to extracting action items — and demonstrate Copilot's core capabilities.

Requirements

• Copilot Chat (free tier works for basic prompts; Copilot Premium unlocks inbox/calendar/document access)
• Microsoft 365 account
• Access via copilot.microsoft.com, Teams, Outlook, or Office apps

Prompts

1. Summarize My Day

Review my calendar and recent activity. Give me a one-paragraph summary of what happened today: key meetings, decisions made, and outstanding items.

2. Draft a Reply

Draft a professional reply to the selected email. Keep it concise, acknowledge the request, provide a clear next step, and match the sender's tone. Offer two versions: one formal, one casual.

3. Explain This Document

Explain the attached document like I'm meeting-pressed and have 2 minutes. Cover: (1) What this is, (2) Why it matters, (3) What action is needed from me.

4. Create a Meeting Agenda

Create a structured agenda for a 60-minute meeting about [topic]. Include: time blocks, discussion items, decision points, and a 10-minute buffer for open discussion. Assign ownership for each agenda item.

5. Find Relevant Files

Search my recent files and emails for everything related to [project/topic]. Return a list of the top 5 most relevant documents with a one-line summary of each and where it's located.

6. Translate This

Translate the selected text into [language]. Preserve the original tone and formatting. After the translation, note any phrases that don't translate directly and suggest culturally appropriate alternatives.

7. Rewrite for Clarity

Rewrite the selected text to be clearer and more concise. Cut filler words, use active voice, shorten sentences, and ensure the main point is in the first two lines. Show the before/after side by side.

8. Extract Action Items

Review the selected meeting transcript or email thread. Extract every action item into a table with columns: Action Item, Owner, Deadline, Status. Flag anything without a clear owner or deadline.

9. Compare Two Documents

Compare the two attached documents. Create a table showing: (1) Areas of agreement, (2) Key differences, (3) Contradictions, (4) Gaps in either document. End with a recommendation on which version to proceed with and why.

10. Give Me a Briefing

Give me a 5-minute briefing on [topic]. Structure it as: (1) Current state, (2) Key developments in the last 30 days, (3) What's at stake, (4) Recommended next steps. Keep it to 5 bullet points max.

Usage Tips

• Start simple: Use prompts 1–3 on day one to see immediate value.
• Select text first: For prompts 2, 3, 7, and 8, highlight the relevant email or text before pasting the prompt.
• Attach files: For prompts 3, 5, and 9, drag files into the Copilot chat window.
• Iterate: Follow up with "Make it shorter", "Add more detail on X", or "Format as a table" to refine output.
• Upgrade path: Prompts 1, 5, and 10 work best with Copilot Premium for full inbox/calendar access.

Generate a structured morning briefing that summarizes your inbox, surfaces action items, and suggests follow-ups from the prior workday.

Requirements

• Copilot Premium or Copilot for Microsoft 365 with Chat enabled
• Access to Outlook inbox and calendar
• Works best in Copilot Chat within Teams, Outlook, or the Copilot web app

Prompt

Review my emails and calendar from the prior workday (yesterday). Generate a structured morning briefing with the following sections:

1. **Executive Summary**: 3–5 sentences summarizing the day's key communications and decisions.
2. **Action Items Table**: List every item where I'm @mentioned or expected to act. Use this table format:

| @mentioned | Topic | Summary | Action Item | Follow-Up |
|---|---|---|---|---|

3. **Open Loops**: Any threads or decisions left unresolved that need attention today.
4. **Today's Priorities**: Based on calendar and inbox urgency, suggest my top 3 priorities for today.

Keep it concise. Use bullet points where appropriate. Flag anything time-sensitive in bold.

Usage Tips

• Best time: Run this within the first 30 minutes of your workday for maximum impact.
• Scope control: Add "Focus only on emails from [specific sender]" or "Include only threads tagged [label]" to narrow scope.
• Frequency: Use daily. Copilot builds context over time, so briefings improve with regular use.
• Customization: Append "Also check my Teams messages" if you want Teams included in the scan.
• Output format: Add "Format as a bulleted briefing I can paste into my daily standup notes" for meeting-ready output.

A three-phase daily structure prompt that keeps you aligned from standup to shutdown. Run it at morning kickoff, midday reset, and end-of-day wrap-up.

Requirements

• Copilot Premium or Copilot for Microsoft 365 with Chat enabled
• Access to Outlook calendar and inbox
• Works in Copilot Chat (Teams, Outlook, Word, or web app)

Prompt

Act as my executive operating rhythm. I need a three-phase daily structure. Generate all three sections below:

---

### PHASE 1: Morning Kickoff (run at start of day)

Review my calendar and inbox for today. Produce:
1. **What's on deck**: List every meeting, deadline, and deliverable scheduled today with times.
2. **Top 3 priorities**: Based on urgency, impact, and dependencies, rank my three most important outcomes for today.
3. **Context primers**: For each priority, summarize the last relevant email or document thread so I'm up to speed.

---

### PHASE 2: Afternoon Reset (run at ~1pm)

1. **Progress check**: Compare my morning priorities against what I've actually completed. Flag any slippage.
2. **Reprioritize**: Based on new emails or calendar changes since morning, adjust my afternoon focus. What should I drop, defer, or double down on?
3. **Blocker scan**: Identify anything stuck waiting on others. Suggest follow-up messages I should send.

---

### PHASE 3: End-of-Day Wrap-Up (run at close of business)

1. **Accomplishments**: Summarize what I shipped, decided, or resolved today in 5 bullet points max.
2. **Open loops**: List unresolved threads, pending decisions, and items carrying to tomorrow.
3. **Tomorrow's preview**: Based on tomorrow's calendar and current open items, preview my top 3 priorities for the next day.
4. **Shutdown note**: One sentence capturing today's overall momentum (green/yellow/red) and why.

---

Keep all sections concise. Use bullet points and tables where helpful. Bold anything time-sensitive.

Usage Tips

• Schedule it: Set calendar reminders at 8:30am, 1:00pm, and 5:00pm to run the corresponding phase.
• Phase isolation: You can paste only one phase at a time. The prompt is structured so each section works independently.
• Integration: Append "Write the wrap-up to my OneNote daily log" if you want automatic journaling.
• Customization: Add "My work hours are [X] to [Y]" to calibrate timing.
• Escalation: Add "If any priority is at risk, draft an escalation message I can send to my manager" for proactive risk management.

Generate a leadership-ready daily status report across all active accounts. Designed for executives who need a single-page view of customer health, risks, and next steps.

Requirements

• Copilot Premium or Copilot for Microsoft 365 with Chat enabled
• Access to Outlook inbox, Teams messages, and relevant SharePoint/OneDrive documents
• Works best in Copilot Chat with attached account files or within Teams

Prompt

Generate a daily executive readout for my active customer accounts. Review my emails, Teams messages, and any attached account documents from the last 24 hours. Produce the following:

---

### Account Status Table

| Customer | Priority | Status | Risk | Next Step |
|---|---|---|---|---|

Populate this table for every account with recent activity. Use Green/Yellow/Red for Status and Low/Medium/High for Risk.

---

### Wins

List any positive developments: deals advanced, milestones hit, stakeholder alignment achieved, or positive feedback received. 3–5 bullets max.

---

### Blockers

List anything stalled, at risk, or requiring escalation. For each blocker, identify:
- What's blocked
- Who owns the resolution
- What I need to do

---

### Escalations

Flag any items requiring immediate leadership attention. For each:
- Issue summary (one line)
- Impact if unresolved
- Recommended action

---

### Today's Focus

Based on the above, recommend my top 3 account-level actions for today, ranked by urgency and impact.

---

Keep it tight. Executives scan this in under 2 minutes. Use bold for anything requiring my direct action.

Usage Tips

• Attach context: Drop account plans, CRM exports, or recent meeting notes into the chat for richer output.
• Scope control: Add "Focus only on [Customer Name] accounts" to narrow the readout.
• Distribution-ready: Append "Format this as an email I can send to my leadership team" for shareable output.
• Cadence: Run daily at 8:00am or 4:00pm depending on your reporting rhythm.
• Historical trend: Add "Compare today's status to last week's readout and note any changes" for continuity.

Review your week like a business case study. This end-of-week self-reflection prompt surfaces patterns, extracts lessons, and generates structured recommendations — all written in third person for objectivity.

Requirements

• Copilot Premium or Copilot for Microsoft 365 with Chat enabled
• Access to Outlook inbox, calendar, and recent documents
• Best run Friday afternoon or Monday morning

Prompt

Review my work communications, calendar, and documents from this past week (last 5 workdays). Analyze the week as a business case study written in third person. Produce the following:

---

### Executive Summary

One paragraph summarizing the week's overall trajectory. What was the central theme or dominant challenge?

---

### What's Working

List 3–5 initiatives, behaviors, or decisions that produced positive outcomes. For each:
- What happened
- Why it worked
- Whether it's repeatable

---

### What's Not Working

List 3–5 areas of friction, missed targets, or recurring problems. For each:
- What happened
- Root cause hypothesis
- Pattern frequency (isolated incident vs. recurring theme)

---

### Structural Analysis

1. **Time allocation**: How was my time actually spent vs. how it should have been spent? Flag misalignments.
2. **Decision quality**: Review key decisions made this week. Which were well-reasoned? Which were reactive or rushed?
3. **Stakeholder dynamics**: Note any relationship patterns — who responded well, where friction occurred, where communication broke down.

---

### Recommendations for Next Week

Based on the analysis above, provide 5 actionable recommendations. Each should include:
- The action
- Expected impact
- Effort level (low/medium/high)

---

### Key Lesson

End with one sentence capturing the single most important lesson from this week.

---

Write in third person throughout ("the executive" not "I"). Keep it analytical, not defensive. Use data from actual communications where possible.

Usage Tips

• Timing: Run Friday at 3pm or Monday at 8am for maximum reflection value.
• Scope: Add "Include Teams messages and meeting transcripts" for broader coverage.
• Continuity: Append "Compare findings to last week's case study and note trends" for longitudinal tracking.
• Actionability: Add "Convert the top 2 recommendations into calendar blocks for Monday" for automatic scheduling.
• Archiving: Save the output to a weekly journal folder for quarterly pattern analysis.

Generate a single-page visual summary of your current workload. This prompt produces a markdown table showing active projects, deadlines, status, and energy level — everything at a glance.

Requirements

• Copilot Premium or Copilot for Microsoft 365 with Chat enabled
• Access to Outlook calendar, inbox, and any project tracking documents
• Works in Copilot Chat (Teams, Outlook, Word, or web app)

Prompt

Generate a single-page "whiteboard" view of my current workload. Review my calendar, open emails, and any attached project documents. Produce a markdown table with these columns:

| Project / Initiative | Owner | Deadline | Status | Energy Level | Next Action |
|---|---|---|---|---|---|

Populate this table for every active project or initiative I'm involved in. Use:
- **Status**: Not Started / In Progress / Blocked / Complete / On Hold
- **Energy Level**: 🔋 (high energy) / 🔌 (moderate) / ⚡ (low — needs attention or delegation)

After the table, add three short sections:

1. **This Week's Focus**: Top 3 items I should prioritize this week.
2. **At Risk**: Anything slipping past deadline or showing red flags.
3. **Delegation Candidates**: Items I could hand off or deprioritize to free capacity.

Keep it to one page. No fluff. This is my daily operating dashboard.

Usage Tips

• Refresh cadence: Run every Monday morning and Wednesday afternoon to keep the board current.
• Attach project files: Drop your project plan, roadmap, or CRM export into the chat for richer data.
• Custom columns: Add "Stakeholders", "Budget", or "Phase" to the column list if your context needs it.
• Export-ready: Append "Format this as a slide I can paste into my weekly leadership update" for presentation output.
• Trend tracking: Save weekly snapshots and compare month-over-month to spot capacity patterns.

A curated pack of 10 high-impact prompts designed specifically for executives and leaders. Each prompt addresses a core leadership challenge — from inbox triage to strategic decision-making.

Requirements

• Copilot Premium or Copilot for Microsoft 365 with Chat enabled
• Access to Outlook, Teams, and relevant documents
• Works in Copilot Chat (Teams, Outlook, Word, or web app)

Prompts

1. Inbox Triage

Review my unread emails from the last 24 hours. Categorize each into: (1) Requires my direct action, (2) Can be delegated, (3) FYI only, (4) Can be deleted. For category 1, draft a suggested response or next step. For category 2, suggest who should own it and draft a delegation message.

2. Meeting Preparation

I have a meeting with [attendees] about [topic] at [time]. Review all emails, documents, and prior meeting notes related to this topic and attendees. Produce: (1) Background brief, (2) Key decisions needed, (3) My talking points, (4) Questions I should ask, (5) Risks to flag.

3. Document Summarization

Summarize the attached document in 5 bullet points capturing: (1) Main thesis or purpose, (2) Key data or findings, (3) Recommendations or calls to action, (4) Any risks or caveats, (5) What requires my attention or decision.

4. Strategic Analysis

Analyze [strategic initiative or decision] from three perspectives: (1) Financial impact over 12 months, (2) Organizational capability required, (3) Competitive positioning. For each perspective, provide: assessment, confidence level, and key assumption.

5. Decision Framing

I need to decide [describe decision]. Frame this as a structured decision memo: (1) Options available, (2) Criteria for evaluation, (3) Recommendation with rationale, (4) What success looks like in 90 days, (5) What could go wrong and mitigation.

6. Risk Assessment

Review my current active projects and communications. Identify the top 5 risks across: (1) Delivery timelines, (2) Stakeholder alignment, (3) Resource constraints, (4) Technical or operational issues, (5) External dependencies. For each risk: likelihood, impact, and recommended mitigation.

7. Stakeholder Management

Map my key stakeholders for [project/initiative]. For each: (1) Their interests and concerns, (2) Current sentiment based on recent communications, (3) What they need from me, (4) Recommended engagement approach. Flag any stakeholder showing negative sentiment.

8. Delegation

Review my current workload. Identify 5 tasks I should delegate. For each: (1) Why it's delegable, (2) Recommended owner, (3) Success criteria, (4) Draft delegation message with clear expectations and deadline.

9. Communication Calibration

Review my last 20 sent emails. Analyze my communication style: (1) Tone (direct vs. diplomatic), (2) Length (concise vs. detailed), (3) Clarity of asks, (4) Response rate from recipients. Provide 3 specific improvements with before/after examples.

10. Weekly Review

Conduct my weekly leadership review. Cover: (1) Decisions made this week and their status, (2) Team performance highlights and concerns, (3) Strategic progress vs. plan, (4) Open issues requiring escalation, (5) Top 3 priorities for next week.

Usage Tips

• Mix and match: These prompts are modular. Use any combination based on your daily leadership challenges.
• Attach context: Drop relevant documents, spreadsheets, or presentations into the chat for richer analysis.
• Cadence: Prompts 1, 2, and 3 are daily; 4–9 are situational; 10 is weekly.
• Customization: Replace bracketed placeholders with your specific context before running.
• Output control: Append "Format as a bulleted briefing" or "Format as an email draft" to match your workflow.

Pressure-test any proposal, decision, or strategy. This prompt surfaces the strongest counterarguments, identifies failure modes, and forces rigorous defense before you commit.

Requirements

• Copilot Premium or Copilot for Microsoft 365 with Chat enabled
• Works with attached documents, pasted text, or verbal descriptions
• Best used before leadership presentations or go/no-go decisions

Prompt

Act as a skeptical senior stakeholder reviewing the following proposal. Pressure-test it rigorously. Do not be polite — be constructive but direct.

[Paste proposal, decision, or strategy here]

Produce the following analysis:

---

### The 3 Most Likely Reasons This Fails

For each failure mode:
1. What goes wrong
2. Why it's likely (cite patterns, data, or precedent)
3. Early warning signs to watch for

---

### The Strongest Case Against This

Write the best opposition argument a smart, informed skeptic would make. Use their likely framing, not yours. Include:
- Core objection
- Supporting evidence or logic
- What they'd propose instead

---

### One Thing Most Likely to Go Wrong

Identify the single highest-probability, highest-impact failure point. Be specific. Then provide:
- The mitigation strategy
- The contingency plan if mitigation fails
- The point of no return (when to kill this)

---

### Questions I Should Be Able to Answer in a Defense

List 5 tough questions a skeptical board member or executive would ask. For each, provide the answer I should give based on available information. Flag any question where I currently lack a credible answer.

---

### Verdict

On a scale of 1–10, how defensible is this proposal? What would move it from current score to an 8 or above?

---

Write in consultant style: direct, structured, evidence-based. No hedging.

Usage Tips

• Pre-meeting prep: Run this 24 hours before any leadership presentation or go/no-go meeting.
• Attach supporting docs: Drop your business case, financial model, or project plan into the chat for grounded analysis.
• Iterate: After reviewing the output, paste your responses to the tough questions and ask "Does this adequately address the concerns?"
• Customization: Add "Focus skepticism on the financial assumptions" or "Focus on execution risk" to target specific dimensions.
• Red team: Run this prompt twice — once as a skeptical CFO, once as a skeptical engineering lead — for multi-lens coverage.

Read any document, email thread, or proposal through the lens of a specific role. This prompt surfaces only what matters to that perspective, filtering out noise and highlighting role-specific concerns.

Requirements

• Copilot Premium or Copilot for Microsoft 365 with Chat enabled
• Works with attached documents, pasted text, or selected content
• Best used for cross-functional review before broad distribution

Prompt

Read the following document through the lens of a [CFO / engineer / customer / board member / HR leader / operations lead]. Surface ONLY what matters to that perspective. Filter out everything else.

[Paste document, email thread, or attach file]

Produce the following:

---

### What This Person Cares About

List the top 3–5 things this role would focus on. For each:
- Why it matters to them
- What they'd want to see
- What's currently missing or unclear

---

### What They'd Question

List every claim, assumption, or recommendation this role would push back on. For each:
- The specific concern
- Why they'd raise it (cite their incentives or constraints)
- What evidence they'd demand

---

### What They'd Ignore

List elements of this document that this role would skip or dismiss. Explain why.

---

### How They'd Rewrite the Recommendation

If this role were editing the final recommendation, what would they change? Provide:
- Their preferred framing
- What they'd add
- What they'd cut

---

### Blind Spots

What is this perspective missing? What would another role see that this one overlooks? Flag 2–3 cross-functional blind spots.

---

Be specific. Reference actual content from the document, not generic advice.

Usage Tips

• Role selection: Replace the bracketed role with any stakeholder type: CFO, CTO, customer success lead, end user, regulator, investor, etc.
• Multi-lens review: Run this prompt 3–4 times with different roles before finalizing any major document.
• Attach the doc: Drag the actual file into Copilot chat for grounded analysis rather than pasting text.
• Pre-flight check: Use this before sending proposals to cross-functional audiences to anticipate objections.
• Customization: Add "This person is risk-averse" or "This person is growth-focused" to calibrate the lens.

Prepare for any upcoming meeting in seconds. This prompt summarizes recent communications with attendees, surfaces prior agreements, and identifies what to push for — so you walk in fully briefed.

Requirements

• Copilot Premium or Copilot for Microsoft 365 with Chat enabled
• Access to Outlook inbox, calendar, and Teams messages
• Works best in Copilot Chat within Outlook or Teams

Prompt

I have an upcoming meeting with [attendees] about [topic] on [date/time]. Brief me thoroughly.

Review all recent emails, Teams messages, calendar notes, and documents involving these attendees and this topic from the last 30 days. Produce:

---

### Background Brief

One paragraph summarizing the history of this topic with these people. What's been discussed, decided, or deferred?

---

### What Was Agreed Last Time

List every commitment, decision, or action item from the last interaction. For each:
- What was agreed
- Who owns it
- Current status (delivered / in progress / overdue)

---

### What to Push For

Based on the context above, recommend:
1. My primary objective for this meeting
2. 2–3 specific asks or decisions I should drive to closure
3. Concessions I can offer if needed

---

### Attendee Intelligence

For each attendee, note:
- Their likely position on this topic
- Any recent sentiment signals (positive / neutral / resistant)
- What they care about that I can leverage

---

### Open Questions

List everything still unresolved that needs to be addressed in this meeting. Prioritize by importance.

---

### Risks & Red Flags

Flag anything that could derail this meeting: misaligned expectations, missing data, stakeholder conflicts, or timing issues.

---

Keep it actionable. I should be able to walk into this meeting fully prepared after reading this.

Usage Tips

• Timing: Run this 2–4 hours before the meeting for optimal prep without information overload.
• Replace placeholders: Fill in [attendees], [topic], and [date/time] before running.
• Attach prior notes: Drop previous meeting notes or agendas into the chat for richer context.
• Rehearsal mode: Append "Now simulate a 5-minute version of this meeting. Play the skeptical attendee and I'll respond" for practice.
• Follow-up automation: After the meeting, paste your notes and ask "Generate follow-up emails based on what was agreed" for instant action items.

Generate a structured monthly business review for your portfolio of accounts. This prompt produces leadership-ready analysis covering performance vs. targets, wins, risks, customer health, pipeline, and actionable recommendations.

Requirements

• Copilot Premium or Copilot for Microsoft 365 with Chat enabled
• Access to Outlook inbox, Teams messages, CRM exports, and account documents
• Best run on the first or last business day of the month

Prompt

Generate a monthly business review for my active customer accounts. Review all communications, documents, and data from the past 30 days. Produce the following structured report:

---

### Performance vs. Targets

For each active account, compare actual performance against monthly targets:
- Revenue / pipeline vs. quota
- Milestones delivered vs. planned
- Key metrics (NPS, CSAT, adoption rate, or other relevant KPIs)

Use a table format: | Account | Target | Actual | Variance | Trend |

---

### Key Wins This Month

List 5–7 significant achievements across all accounts:
- Deals closed or advanced
- Milestones delivered
- New stakeholder relationships established
- Positive feedback or expansion signals

For each win, note the account and why it matters.

---

### Risks & Issues

Identify every account or initiative showing risk signals:
- Churn risk indicators
- Escalations or complaints
- Delivery slippage
- Stakeholder friction

For each risk: severity (low/medium/high), root cause, and recommended mitigation.

---

### Customer Health Summary

Provide a health score for each active account (1–10) based on:
- Engagement level (meeting frequency, response rates)
- Sentiment (tone of communications, feedback)
- Delivery status (on track / at risk / behind)
- Expansion signals (upsell/cross-sell opportunities)

Flag any account scoring below 5 with a one-line explanation.

---

### Pipeline Outlook

Summarize the next 30-day pipeline:
- Opportunities in active pursuit
- Expected close dates and probabilities
- Key decisions needed to advance deals
- Accounts requiring executive sponsorship

---

### Recommendations for Next Month

Provide 5–7 actionable recommendations ranked by impact:
1. What to double down on
2. What to fix immediately
3. What to deprioritize or hand off
4. Strategic bets worth making

For each: expected outcome and effort required.

---

### One-Page Executive Summary

End with a 5-bullet summary I can paste into a leadership email:
- Overall portfolio health
- Top win
- Biggest risk
- Key ask from leadership
- Next month's focus

---

Keep it data-driven. Reference actual communications and documents where possible.

Usage Tips

• Attach data: Drop CRM exports, revenue reports, or account scorecards into the chat for grounded analysis.
• Scope control: Add "Focus only on strategic accounts" or "Include all mid-market accounts" to narrow or broaden.
• Quarterly extension: Append "Compare to last quarter's review and note trends" for quarterly business reviews.
• Distribution-ready: Add "Format this as a slide deck outline" for leadership presentation output.
• Cadence: Run monthly; save outputs to build a historical review archive for trend analysis.

Simulate a council of advisors reviewing a critical decision. Each advisor represents a distinct lens — strategist, skeptic, operator, communicator — and provides an independent assessment before the council converges on a recommendation.

Requirements

• Copilot Premium or Copilot for Microsoft 365 with Chat enabled
• Works with attached documents, pasted proposals, or verbal descriptions
• Best used for high-stakes decisions, strategy reviews, or pre-commitment validation

Prompt

Simulate a Guardian Council reviewing the following decision. Four advisors each provide an independent assessment from their perspective, then the council converges on a unified recommendation.

[Describe the decision, proposal, or strategy here]

---

### Advisor 1: The Strategist

**Lens**: Long-term positioning, competitive advantage, alignment with objectives

Assess:
1. Does this move us toward our strategic goals?
2. What's the 12-month trajectory if we proceed?
3. What strategic opportunities does this create or foreclose?
4. Is this the best use of our strategic capital right now?

Provide a clear verdict: STRATEGIC YES / STRATEGIC NO / STRATEGIC MAYBE with conditions.

---

### Advisor 2: The Skeptic

**Lens**: Risk, failure modes, hidden costs, unintended consequences

Assess:
1. What are the 3 most likely ways this fails?
2. What costs or trade-offs aren't being acknowledged?
3. What would a smart opponent do in response?
4. What's the worst-case scenario and can we survive it?

Provide a clear verdict: GO / NO-GO / GO WITH MITIGATIONS (list them).

---

### Advisor 3: The Operator

**Lens**: Execution feasibility, resource requirements, timeline realism, team capacity

Assess:
1. Do we have the people, skills, and bandwidth to execute this?
2. Is the timeline realistic given current commitments?
3. What operational dependencies exist and are they secure?
4. What's the minimum viable version we could ship in half the time?

Provide a clear verdict: EXECUTABLE / EXECUTABLE WITH CHANGES / NOT YET EXECUTABLE.

---

### Advisor 4: The Communicator

**Lens**: Stakeholder buy-in, narrative strength, change management, messaging risk

Assess:
1. How will key stakeholders react to this? Who supports, who resists?
2. Is the story compelling and defensible?
3. What's the change management burden?
4. What messaging landmines should we avoid?

Provide a clear verdict: COMMUNICABLE / COMMUNICABLE WITH PREP / HIGH COMMUNICATION RISK.

---

### Council Deliberation

Now synthesize all four perspectives. Address:
1. Where do the advisors agree?
2. Where do they conflict, and which perspective carries more weight?
3. What conditions must be met before proceeding?
4. What's the single biggest risk across all lenses?

---

### Final Recommendation

The council issues one of three verdicts:

- **PROCEED**: All advisors support or can be satisfied with mitigations
- **PROCEED WITH CONDITIONS**: List the non-negotiable conditions
- **HOLD**: The decision isn't ready; specify what's missing

Include a one-paragraph rationale the council can stand behind.

---

Write each advisor's section in their voice: Strategist is visionary, Skeptic is sharp, Operator is practical, Communicator is diplomatic. The deliberation and recommendation should be unified and decisive.

Usage Tips

• High-stakes decisions: Use this before go/no-go meetings, major investments, or strategic pivots.
• Attach context: Drop the business case, financial model, or project plan into the chat for grounded assessments.
• Custom advisors: Replace any advisor with a domain-specific lens (e.g., "The CFO", "The Security Lead", "The Customer Advocate").
• Pre-meeting prep: Run this 24 hours before leadership presentations to anticipate objections.
• Documentation: Save the council output as a decision record for audit and future reference.

Senior copywriter-level editing. This prompt proofreads your text, trims fat, improves clarity, corrects errors, shifts to active voice, shortens sentences, and delivers an informal-yet-professional tone — with every change explained.

Requirements

• Copilot Chat (free tier sufficient; no inbox/calendar access needed)
• Works in any Copilot surface: Word, Outlook, Teams, or web app
• Paste text or select content before running

Prompt

Act as a senior copywriter. Edit the following text with precision. Apply these rules in order:

1. **Proofread**: Fix all grammar, spelling, punctuation, and factual inconsistencies.
2. **Trim fat**: Remove filler words, redundant phrases, and unnecessary qualifiers.
3. **Improve clarity**: Ensure the main point is in the first two sentences. Cut ambiguity.
4. **Active voice**: Convert passive constructions to active wherever possible.
5. **Shorten sentences**: Break any sentence over 25 words into two. Target average sentence length under 18 words.
6. **Tone**: Informal yet professional. Confident but not arrogant. Direct but not blunt.

After delivering the edited version, provide a change log:

| Original | Edited | Reason |
|---|---|---|

List every substantive change (not just typo fixes). Explain why each edit improves the text.

[Paste text to edit here]

Usage Tips

• Email drafting: Paste your draft email before sending. This prompt catches tone misfires and clarity issues.
• Document editing: In Word, select the passage and run this in Copilot in Word for inline editing.
• Batch editing: For long documents, paste one section at a time (500–800 words max) for best results.
• Tone calibration: Add "Make it more formal" or "Make it more casual" to shift the tone dial.
• Before/after comparison: Append "Show the original and edited version side by side" for quick visual review.

Mine a year of your communications into a reusable leadership voice profile. This prompt extracts your patterns — word choice, sentence structure, tone markers — and produces a style guide you can reference for consistent, authentic communication.

Requirements

• Copilot Premium or Copilot for Microsoft 365 with Chat enabled
• Access to Outlook sent items (last 6–12 months of emails)
• Works best in Copilot Chat with attached email exports or within Outlook

Prompt

Analyze my sent emails from the past [6/12] months. Mine them to create a reusable Leadership Voice Profile. Extract the following patterns:

---

### Word Choice & Vocabulary

1. **Signature phrases**: What words or phrases do I use repeatedly? List the top 10.
2. **Jargon level**: Do I use industry terminology, plain language, or a mix? Rate 1–10.
3. **Formality markers**: What words signal formality vs. casualness in my writing?
4. **Words to keep**: Which phrases authentically represent my voice and should be preserved?
5. **Words to drop**: Which phrases are overused, vague, or undermine my credibility?

---

### Sentence Structure

1. **Average sentence length**: What's my typical sentence length?
2. **Paragraph structure**: Do I lead with conclusions, build to them, or bury the lead?
3. **List usage**: How often do I use bullet points, numbered lists, or tables?
4. **Rhythm**: Is my writing staccato (short, punchy) or flowing (longer, connected)?

---

### Tone Markers

1. **Directness**: Rate my directness 1–10. Do I get to the point fast or build context first?
2. **Warmth vs. efficiency**: Where do I fall on this spectrum? Provide examples.
3. **Humor**: Do I use humor? If so, what kind (dry, self-deprecating, witty)?
4. **Emotional signaling**: Do I express enthusiasm, concern, or neutrality? How frequently?
5. **Call-to-action style**: How do I make asks? Direct commands, questions, or soft requests?

---

### What to Keep

List 5 elements of my current communication style that are working well and should be preserved.

---

### What to Drop

List 5 habits that undermine clarity, credibility, or efficiency. For each:
- The habit
- Why it's a problem
- What to do instead

---

### Voice Profile Summary

End with a 100-word "voice brief" I can paste into future prompts to calibrate Copilot to write in my style. Include:
- Tone
- Structure preferences
- Signature elements
- What to avoid

---

Be honest, not flattering. This is a diagnostic, not a compliment session.

Usage Tips

• Time window: Adjust the [6/12] month range based on how much history you want analyzed.
• Narrow scope: Add "Focus only emails to my direct reports" or "Focus only on executive-level communications" for role-specific profiles.
• Reuse the output: Save the "Voice Profile Summary" and paste it into future Copilot prompts with "Write this in my leadership voice: [paste profile]".
• Quarterly refresh: Re-run every quarter to track how your voice evolves.
• Comparison mode: Add "Compare my voice to [colleague's name]'s based on their emails" for relative analysis.

Turn meeting transcripts, notes, or email threads into a structured decisions-and-owners table. Extract every commitment, assign ownership, set deadlines, and track dependencies.

Requirements

• Copilot Premium or Copilot for Microsoft 365 with Chat enabled
• Works with Teams meeting transcripts, Outlook email threads, or pasted notes
• Best used immediately after meetings while context is fresh

Prompt

Review the following meeting transcript/notes. Extract every decision, action item, and commitment into a structured table.

[Paste transcript, notes, or attach file]

Produce the following:

---

### Decisions & Action Items Table

| Decision / Action Item | Owner | Deadline | Status | Dependencies |
|---|---|---|---|---|

Populate this table for every explicit or implicit commitment made during the meeting. Use:
- **Status**: Not Started / In Progress / Complete / Blocked / Deferred
- **Dependencies**: List any person, deliverable, or decision this item depends on. Put "None" if standalone.

---

### Key Decisions Made

List every decision finalized during the meeting. For each:
- The decision
- Who made it
- Rationale (if stated)
- Any dissent or concerns raised

---

### Open Questions

List every question raised but not answered. For each:
- The question
- Who needs to answer it
- Why it matters (impact if unresolved)

---

### Follow-Up Required

Identify anything that needs a follow-up meeting, deeper analysis, or escalation. For each:
- What needs follow-up
- Recommended next step
- Suggested timeline

---

### Draft Follow-Up Email

Generate a follow-up email I can send to all attendees summarizing decisions, action items, and next steps. Keep it under 150 words. Include the table above.

---

Be precise. If an owner isn't explicitly named, flag it as "UNASSIGNED — needs owner." If no deadline was set, flag it as "TBD."

Usage Tips

• Teams integration: In Teams, open the meeting transcript and paste this prompt directly into Copilot in Teams for instant extraction.
• Multi-meeting batch: Paste notes from 2–3 meetings and add "Process each meeting separately with clear section breaks" for batch processing.
• Status tracking: Re-run this prompt weekly with updated notes and add "Compare to last week's action items table and note progress" for continuity.
• Owner resolution: Append "Look up the org chart and suggest likely owners for any UNASSIGNED items" for automatic owner inference.
• Distribution: The draft follow-up email is ready to copy-paste into Outlook. Review names and deadlines before sending.

10 high-impact prompts for IT leaders covering deployment status, adoption metrics, license optimization, governance, user feedback, ROI, training needs, risk assessment, vendor comparison, and roadmap planning.

Requirements

• Copilot Premium or Copilot for Microsoft 365 with Chat enabled
• Access to relevant documents, reports, and communications
• Works in Copilot Chat (Teams, Outlook, Word, or web app)

Prompts

1. Deployment Status Dashboard

Review all IT deployment-related communications and documents from the last 30 days. Generate a deployment status dashboard with:

| Initiative | Phase | Start Date | Target Date | Current Status | Risk Level | Blockers |
|---|---|---|---|---|---|---|

Populate for every active IT deployment. Flag anything behind schedule or showing high risk. End with 3 recommended actions to unblock stalled deployments.

2. Adoption Metrics Analysis

Analyze user adoption data for [tool/platform]. Review feedback emails, support tickets, and usage reports. Produce:
1. Current adoption rate vs. target
2. Top 3 adoption barriers identified from user feedback
3. Departments with highest and lowest adoption
4. Recommended interventions ranked by expected impact
5. Draft communication to low-adoption departments explaining value and offering support

3. License Optimization

Review our current software license inventory and usage data. Identify:
1. Underutilized licenses (usage < 20% over last 90 days)
2. Over-licensed users (premium licenses with basic usage patterns)
3. Duplicate tooling (multiple tools serving same function)
4. Upcoming renewals and cost projections

Produce a table: | Tool | Current Licenses | Active Users | Utilization % | Monthly Cost | Recommendation |

End with total estimated savings from recommended changes.

4. Governance Checklist

Generate an IT governance audit checklist for [system/initiative]. Cover:
1. Access controls and least privilege compliance
2. Data classification and handling procedures
3. Change management process adherence
4. Backup and disaster recovery status
5. Security patch currency
6. Vendor SLA compliance

For each item: current status (compliant/non-compliant/partial), evidence, and remediation steps if non-compliant.

5. User Feedback Analysis

Review all IT-related user feedback from the last 30 days (emails, tickets, surveys, Teams messages). Produce:
1. Sentiment breakdown (positive/neutral/negative %)
2. Top 5 recurring themes with examples
3. Urgent issues requiring immediate attention
4. Feature requests grouped by category
5. Draft response to the top 3 complaints with resolution timeline

6. ROI Calculation

Calculate the ROI for [IT initiative/tool]. Review available data on:
1. Implementation costs (licenses, labor, training)
2. Time saved or efficiency gains
3. Error reduction or quality improvements
4. Revenue impact (if applicable)
5. Ongoing maintenance costs

Produce: total investment, monthly benefit, payback period, 12-month projected ROI %, and 3 assumptions that could change the calculation.

7. Training Needs Assessment

Based on user feedback, support tickets, and adoption data, identify training gaps for [tool/platform]. Produce:
1. Skills gap analysis by department
2. Recommended training programs (self-paced vs. instructor-led)
3. Priority ranking based on impact and audience size
4. Draft training schedule for next quarter
5. Success metrics to track training effectiveness

8. Risk Assessment

Conduct an IT risk assessment for our current environment. Review incident reports, security alerts, and infrastructure data. Produce:

| Risk Category | Specific Risk | Likelihood | Impact | Current Mitigation | Gap |
|---|---|---|---|---|---|

Cover: cybersecurity, infrastructure, compliance, vendor dependency, and operational risks. End with top 3 risks requiring immediate board attention.

9. Vendor Comparison

Compare [Vendor A] vs [Vendor B] for [use case]. Analyze:
1. Feature parity matrix
2. Total cost of ownership over 3 years
3. Integration complexity with our current stack
4. Security and compliance posture
5. Customer references and market reputation

Produce a scored comparison table and a final recommendation with rationale. Flag any deal-breakers.

10. Roadmap Planning

Based on current IT initiatives, strategic goals, and resource constraints, draft a 6-month IT roadmap. Include:
1. Must-deliver initiatives (aligned to business objectives)
2. Nice-to-have improvements
3. Resource allocation by quarter
4. Dependency mapping between initiatives
5. Risk mitigation for each major milestone

Format as a phased plan: Q1 immediate actions, Q2–Q3 scaling, Q4 optimization.

Usage Tips

• Attach data: Drop license reports, ticket exports, or infrastructure dashboards into the chat for grounded analysis.
• Customize placeholders: Replace bracketed text with your specific tools, platforms, or initiatives.
• Cadence: Prompts 1, 2, 5 are weekly; 3, 4, 8 are monthly; 6, 7, 9, 10 are quarterly or situational.
• Executive output: Append "Format this as a board-ready slide" for leadership presentation output.
• Cross-reference: Run prompts 1 and 8 together for a complete operational health picture.

A focused prompt pack for Copilot in Outlook. Covers inbox triage, long-thread summaries, reply drafting with tone control, meeting prep, calendar optimization, and email analytics.

Requirements

• Copilot Premium or Copilot for Microsoft 365 with Chat enabled
• Outlook desktop, web, or mobile with Copilot button visible
• Works best in Copilot in Outlook or Copilot Chat with Outlook context

Prompts

1. Inbox Triage

Review my unread Outlook inbox. Categorize every email into:
1. **Action required** (needs my direct response or decision)
2. **Delegate** (can be handled by someone else)
3. **FYI** (no action needed, just awareness)
4. **Noise** (can be deleted or archived)

For category 1: draft a suggested reply or next step.
For category 2: suggest the owner and draft a delegation message.
For category 3: provide a one-line summary.

Present as a table: | Priority | From | Subject | Category | Suggested Action |

2. Long-Thread Summary

Summarize this email thread in 5 bullet points:
1. What is the core issue or request?
2. What has been decided so far?
3. What is still open or unresolved?
4. Who are the key stakeholders and their positions?
5. What action is needed from me?

After the summary, draft a concise reply that moves the thread to closure.

3. Reply Drafting with Tone Control

Draft a reply to this email. Provide three versions:
1. **Direct**: Get straight to the point, minimal pleasantries
2. **Diplomatic**: Acknowledge context, soften asks, preserve relationships
3. **Executive**: Brief, decision-focused, suitable for senior leadership

For each version, note when it's appropriate to use.

4. Meeting Prep from Calendar

Review my next 3 calendar meetings. For each, produce:
1. Meeting purpose and expected outcomes
2. Key attendees and their roles
3. Relevant emails or documents from the last 7 days involving these people or topics
4. One preparation question I should have an answer for
5. One risk to flag during the meeting

Keep each briefing to 5 lines max.

5. Calendar Optimization

Analyze my calendar for the next week. Identify:
1. **Back-to-back meetings** with no buffer — suggest where to add 15-minute breaks
2. **Low-value meetings** I could decline or send a delegate to — explain why
3. **Deep work blocks** — identify available focus time and recommend what to schedule there
4. **Meeting stack opportunities** — suggest which meetings could be combined
5. **Time zone conflicts** — flag any meetings at inconvenient hours

Present findings in a table with recommendations.

6. Email Analytics & Communication Review

Analyze my sent emails from the last 30 days. Produce:
1. **Volume**: Total emails sent, average per day, peak sending hours
2. **Response patterns**: Average response time from recipients, who responds fastest/slowest
3. **Thread length**: Average exchanges per conversation — flag any thread over 10 messages
4. **Tone analysis**: Breakdown of formal vs. informal, direct vs. diplomatic
5. **Top recipients**: Who I email most frequently and why
6. **Improvement recommendations**: 3 specific changes to reduce email volume or improve effectiveness

Usage Tips

• Outlook integration: Click the Copilot button in Outlook and paste these prompts directly for context-aware results.
• Select first: For prompts 2 and 3, highlight the email thread or message before running.
• Daily cadence: Run prompt 1 (inbox triage) each morning; prompt 6 (analytics) weekly.
• Calendar hygiene: Run prompt 5 (optimization) every Monday to structure your week proactively.
• Tone library: Save your preferred tone version from prompt 3 and reuse it as a template for future replies.

Generate a curated daily briefing on AI industry developments. This prompt summarizes top news, product updates, regulatory changes, and market movements in a structured format with sources.

Requirements

• Copilot Premium or Copilot for Microsoft 365 with Chat enabled
• Web search capability enabled in Copilot settings
• Works in Copilot Chat (Teams, Outlook, or web app)

Prompt

Generate today's AI industry briefing. Search for the latest AI news from the past 24 hours and produce a structured briefing with the following sections:

---

### Top Stories

List the 5 most significant AI developments today. For each:
- Headline
- One-paragraph summary
- Why it matters (business impact, technological significance, or market implication)
- Source link

---

### Product & Platform Updates

Cover any new AI product launches, feature releases, or platform updates from major vendors (Microsoft, Google, OpenAI, Anthropic, Meta, Amazon, etc.). For each:
- What changed
- Who it affects
- Adoption implications

---

### Regulatory & Policy Changes

List any new AI regulations, compliance requirements, or policy developments. For each:
- Jurisdiction
- What changed
- Deadline or effective date
- Impact on enterprises

---

### Market Movements

Summarize AI-related market activity:
- Stock movements for major AI companies
- M&A activity
- Funding rounds or IPOs
- Analyst commentary

---

### Research & Breakthroughs

Note any significant AI research papers, model releases, or technical breakthroughs. For each:
- What was achieved
- Practical applications
- Timeframe to production relevance

---

### What to Watch Tomorrow

End with 3 items to monitor in the next 24–48 hours: upcoming announcements, earnings calls, regulatory deadlines, or scheduled releases.

---

Include source links for every claim. Keep the total briefing scannable in under 3 minutes.

Usage Tips

• Timing: Run this first thing in the morning for the day's briefing.
• Web search required: Ensure Copilot's web search is enabled in settings for live results.
• Scope control: Add "Focus only on enterprise AI" or "Include generative AI only" to narrow the briefing.
• Distribution-ready: Append "Format this as an email I can send to my leadership team" for shareable output.
• Customization: Add "Also cover AI developments in [specific industry]" for sector-specific briefings.

Configure Copilot to match your role, communication style, priorities, and working context. This prompt produces a personalization profile you can reuse to calibrate every future Copilot interaction.

Requirements

• Copilot Premium or Copilot for Microsoft 365 with Chat enabled
• Access to your Outlook sent items, calendar, and recent documents
• Works in Copilot Chat (Teams, Outlook, or web app)

Prompt

Help me configure Copilot to match my role and working style. Answer the following questions by analyzing my recent communications, calendar patterns, and documents. Then produce a reusable personalization profile.

---

### Role Definition

Based on my calendar meetings, email recipients, and document topics, what is my primary role? What secondary responsibilities do I handle? Define my role profile in 3–5 sentences.

---

### Communication Style

Analyze my last 30 sent emails. Extract:
1. **Formality level**: Rate 1–10 with examples
2. **Directness**: Do I lead with conclusions or build context?
3. **Preferred length**: Am I concise or detailed?
4. **Signature phrases**: What words or structures do I use repeatedly?
5. **Tone markers**: Warm, neutral, or direct?

---

### Priority Framework

Based on my calendar and email patterns, what are my recurring priority categories? For example:
- Customer/account management
- Internal leadership/strategy
- Project delivery
- Stakeholder communication

Rank them by time allocation and flag any misalignment between where my time goes and where it should go.

---

### Working Context

1. **Team size**: How many direct reports or frequent collaborators do I interact with?
2. **Meeting load**: Average meetings per day, total meeting hours per week
3. **Deep work windows**: When do I typically have focus time?
4. **Time zones**: What regions do I regularly work across?

---

### Personalization Profile Output

Produce a compact profile I can paste into future Copilot prompts:

```
ROLE: [role definition]
STYLE: [communication style summary]
PRIORITIES: [top 3 priority categories]
CONTEXT: [team size, meeting load, time zones]
PREFERENCE: [output format preference — tables, bullets, briefings]
```

---

### Memory Configuration Recommendations

Based on the analysis, recommend:
1. What Copilot should remember about me (role, preferences, context)
2. What NOT to save (sensitive data, transient info)
3. How often to refresh this profile

Usage Tips

• Run once, reuse always: Save the output profile and paste it at the start of new Copilot sessions for consistent behavior.
• Quarterly refresh: Re-run every quarter or after a role change to keep the profile current.
• Scope control: Add "Focus only on emails from the last 90 days" to narrow the analysis window.
• Team adaptation: Run this for each direct report to create team-wide personalization profiles.
• Integration: Append "Write this profile to a OneNote page I can reference" for persistent storage.

A practical guide to configuring Copilot's memory. What's worth having Copilot remember, what improves future interactions, and what should never be saved.

Requirements

• Copilot Premium or Copilot for Microsoft 365 with Chat enabled
• Memory settings accessible via Copilot settings or profile configuration
• Works in Copilot Chat (Teams, Outlook, or web app)

Prompt

Help me configure what Copilot should remember about me. Review my recent interactions and communications, then produce a memory configuration guide with three sections:

---

### WHAT TO SAVE — Durable Facts & Preferences

Recommend Copilot to remember these durable, reusable items:

1. **Role & Context**: My job title, team, reporting structure, and primary responsibilities
2. **Communication Preferences**: Preferred tone, format (tables vs. bullets), level of detail, and output length
3. **Recurring Priorities**: My top 3–5 ongoing focus areas that appear across meetings and emails
4. **Key Relationships**: Frequent collaborators, stakeholders, and their roles
5. **Working Patterns**: My typical work hours, meeting-heavy days, and deep work windows
6. **Decision Framework**: How I prefer decisions framed (options + recommendation, pros/cons, executive summary first)
7. **Project Context**: Active initiatives, their status, and my role in each

For each item, provide the exact phrasing I should use when telling Copilot to remember it.

---

### WHAT NOT TO SAVE — Sensitive & Transient Data

Explicitly tell Copilot to NOT remember:

1. **Personal Identifiable Information**: Home addresses, personal phone numbers, family details
2. **Financial Data**: Salaries, compensation details, personal financial information
3. **Health Information**: Medical conditions, leave details, wellness program participation
4. **Confidential HR Data**: Performance reviews, disciplinary actions, promotion candidates
5. **Credentials**: Passwords, API keys, access codes, security questions
6. **Transient Context**: One-off meeting details, daily standup notes, temporary troubleshooting steps
7. **Speculative Content**: Unverified rumors, gossip, or unconfirmed organizational changes

For each category, explain the risk of saving it.

---

### MEMORY MAINTENANCE

1. **Review cadence**: How often should I audit what Copilot remembers? (Recommend quarterly)
2. **Update process**: How to correct outdated memories or add new preferences
3. **Reset procedure**: How to clear Copilot's memory if needed
4. **Scope control**: How to limit what Copilot accesses in different contexts (e.g., personal vs. work chats)

---

End with a one-paragraph "memory policy" I can save and reference.

Usage Tips

• Run at onboarding: Use this when first setting up Copilot to establish clean memory boundaries.
• Quarterly audit: Re-run every quarter to review and update saved memories.
• Explicit commands: Use "Remember that..." and "Forget..." commands in Copilot to manage memory directly.
• Context separation: Keep personal and work Copilot chats separate to prevent memory bleed.
• Compliance: In regulated industries, consult your legal team before enabling broad memory retention.

Compare your current skills to your target role, identify gaps, and generate a structured learning plan with resources. Designed for professionals preparing for promotion, career transitions, or upskilling.

Requirements

• Copilot Premium or Copilot for Microsoft 365 with Chat enabled
• Works with attached resume, performance reviews, or self-assessment documents
• Best used during annual review cycles or career planning sessions

Prompt

Conduct a skill gap analysis and growth plan for my career progression. Use the following inputs:

**Current Role**: [describe your current role, title, responsibilities]
**Target Role**: [describe your target role, title, desired responsibilities]
**Timeframe**: [e.g., 6 months, 12 months, 18 months]

[Paste resume, performance review, or attach relevant documents]

Produce the following:

---

### Current Skills Inventory

Based on my inputs, list my demonstrated competencies grouped by category:
1. **Technical/Hard Skills**: Tools, platforms, methodologies
2. **Leadership/Soft Skills**: Communication, influence, decision-making
3. **Domain Knowledge**: Industry expertise, functional knowledge
4. **Scale & Scope**: Team size, budget, geographic reach, complexity

Rate each skill: Proficient / Developing / Not Yet Demonstrated

---

### Target Role Requirements

What does the target role demand? List required competencies in the same four categories. For each:
- Why it matters for the target role
- Evidence of demand (industry standards, job descriptions, market trends)

---

### Gap Analysis

| Skill Area | Current Level | Target Level | Gap Severity | Priority |
|---|---|---|---|---|

Populate for every gap identified. Use:
- **Gap Severity**: Minor / Moderate / Critical
- **Priority**: Immediate (next 90 days) / Medium-term (3–6 months) / Long-term (6–12 months)

---

### Learning Plan

For each high-priority gap, create a learning action item:
1. **What to learn**: Specific skill or knowledge area
2. **How to learn**: Recommended resources (courses, books, certifications, on-the-job projects)
3. **Time investment**: Estimated hours or weeks
4. **Success metric**: How I'll know I've closed the gap
5. **Accountability**: Who can validate my progress (manager, mentor, peer)

---

### Quick Wins

List 3–5 gaps I can close in under 30 days with high visibility. These should demonstrate progress toward the target role immediately.

---

### Risks to Address

What could derail this plan? Identify:
1. Time constraints and how to mitigate them
2. Resource gaps (budget, access, sponsorship)
3. Organizational barriers (politics, timing, headcount)
4. Personal risks (burnout, scope creep, loss of momentum)

For each risk: mitigation strategy.

---

### 90-Day Checkpoint

Define what success looks like at 90 days:
- Skills demonstrated
- Projects completed
- Conversations had with manager
- Evidence of progression toward target role

---

Be specific and actionable. This is a real plan, not aspirational fluff.

Usage Tips

• Attach your resume: Drop your current resume or LinkedIn profile into the chat for grounded analysis.
• Realistic timeframes: Adjust the [Timeframe] placeholder based on your actual promotion cycle or career goal.
• Manager alignment: Share the output with your manager to align on development priorities.
• Quarterly check-ins: Re-run at 90-day intervals with updated progress to track closure of gaps.
• Customization: Add "Focus on leadership skills" or "Focus on technical upskilling" to narrow the analysis.

Reframe challenges, celebrate wins, and maintain momentum without toxic positivity. These prompts deliver honest, constructive self-assessment that builds resilience and forward motion.

Requirements

• Copilot Premium or Copilot for Microsoft 365 with Chat enabled
• Access to Outlook inbox, calendar, and recent documents
• Works in Copilot Chat (Teams, Outlook, or web app)

Prompts

1. Reframe This Challenge

I'm facing this challenge: [describe the situation]

Reframe it constructively. Do not minimize the difficulty or offer empty optimism. Instead:
1. Name what's actually hard about this (validate the reality)
2. Identify 2–3 strengths or resources I already have that apply
3. Reframe the obstacle as a specific problem to solve (not a personal failing)
4. Suggest one small action I can take in the next 2 hours
5. End with one sentence capturing what this challenge is building in me long-term

Keep it real. No toxic positivity.

2. Celebrate This Week's Wins

Review my communications, calendar, and documents from this past week. Identify every win — big and small — that I may be undercounting.

List them in three categories:
1. **Deliverables shipped**: What did I actually complete?
2. **Relationships strengthened**: Who responded well, aligned, or gave positive feedback?
3. **Personal growth**: What did I handle better than I would have 6 months ago?

For each win, write one sentence on why it matters. End with: "You're not behind. Here's the evidence."

3. Maintain Momentum

I feel my momentum slipping. Review my last 2 weeks of activity and produce:
1. **Progress proof**: List 5 things I've moved forward, even partially
2. **Pattern recognition**: What's working that I should double down on?
3. **Friction points**: What's draining energy that I can eliminate or delegate?
4. **Reset plan**: 3 specific actions for tomorrow that rebuild momentum
5. **Perspective check**: One paragraph reminding me why this work matters

Be direct. I don't need cheerleading — I need a course correction.

4. Constructive Self-Critique

Review my recent work (emails, documents, meeting performance based on available data). Provide an honest assessment:

1. **What I'm doing well**: 3 specific strengths with evidence
2. **Where I'm underperforming**: 3 specific areas with evidence
3. **Blind spots**: What am I not seeing that others likely are?
4. **The fix**: One actionable change for each underperformance area

Balance is critical. Don't inflate strengths or exaggerate weaknesses. This is a diagnostic, not a roast or a compliment session.

5. Pre-Meeting Confidence Calibration

I have an important meeting about [topic] with [attendees]. I'm feeling [nervous/underprepared/overconfident].

Calibrate my mindset:
1. **What I bring to the room**: List my relevant expertise, data, or perspective
2. **What I don't know**: Identify gaps honestly and suggest how to handle them if asked
3. **Likely objections**: What will pushback look like and how to respond
4. **Success definition**: What does a good outcome look like? (It's not perfection)
5. **One-line anchor**: A sentence I can repeat before walking in

Keep it grounded. Confidence comes from preparation, not affirmation.

Usage Tips

• Timing matters: Run "Celebrate Wins" on Friday afternoon; "Maintain Momentum" mid-week when energy dips.
• Honesty over comfort: These prompts are designed to be constructive, not flattering. Trust the output even when it's uncomfortable.
• Customize the challenge: Replace bracketed text with your specific situation for grounded reframing.
• Pair with action: Always follow a positivity prompt with a concrete next step — motivation without execution is just entertainment.
• Weekly rhythm: Run "Constructive Self-Critique" once weekly to maintain honest self-assessment without spiraling.

Identify the 3 traits that could hold you back professionally, in content creation, and in collaboration. A consultant-style blind spot breakdown with actionable remediation advice.

Requirements

• Copilot Premium or Copilot for Microsoft 365 with Chat enabled
• Access to Outlook sent items, Teams messages, and recent documents
• Works in Copilot Chat (Teams, Outlook, or web app)

Prompt

Conduct a blind spot analysis of my communication and work patterns. Review my sent emails, Teams messages, meeting notes, and documents from the last 90 days. Identify the 3 traits most likely to hold me back in three domains:

---

### Domain 1: Professional Advancement

What trait is most likely to limit my career progression? Base this on:
- How I communicate upward (to executives, sponsors, board)
- How I handle visibility and self-promotion
- How I respond to feedback or criticism
- Patterns in my decision-making under pressure

For this trait:
1. Name it directly
2. Show evidence from my actual communications
3. Explain the career risk if unaddressed
4. Provide 3 specific behavioral changes to mitigate it

---

### Domain 2: Content Creation & Communication

What trait undermines the quality, clarity, or impact of what I produce? Base this on:
- Email length and structure patterns
- Document clarity and organization
- Meeting communication effectiveness
- Tone consistency across audiences

For this trait:
1. Name it directly
2. Show evidence from my actual writing
3. Explain the impact on my audience
4. Provide 3 specific improvements with before/after examples

---

### Domain 3: Collaboration & Team Dynamics

What trait creates friction in how I work with others? Base this on:
- How I delegate (or fail to)
- How I handle conflict or disagreement
- How I respond to others' ideas (build on them or shoot them down)
- Meeting behavior (dominating, withdrawing, facilitating)

For this trait:
1. Name it directly
2. Show evidence from team interactions
3. Explain the collaboration cost
4. Provide 3 specific relationship behaviors to adopt

---

### Cross-Domain Pattern

Is there one underlying pattern connecting all three traits? If so, name it and explain the root cause.

---

### 90-Day Remediation Plan

For each trait, define:
1. One habit to start
2. One habit to stop
3. One person to ask for feedback on this specific trait
4. One metric to track improvement

---

Write in consultant style: direct, evidence-based, structured. No hedging, no sugar-coating. This is a diagnostic.

Usage Tips

• Timing: Run annually or during performance review cycles for maximum impact.
• Feedback loop: Share the output with your manager or coach and ask "Do you see these patterns too?"
• Evidence grounding: The analysis is only as good as your communication history. Ensure 90+ days of data is accessible.
• Follow-up: Re-run after 90 days with "Compare to my previous blind spot analysis and assess progress" for tracking.
• Psychological safety: This prompt will produce uncomfortable output. That's the point. Sit with it before reacting.

Understand what crosses the tenant boundary when web search is enabled, BAA coverage scope, and regulated-industry guidance for using Copilot with sensitive data.

Requirements

• Copilot Premium or Copilot for Microsoft 365 with Chat enabled
• Applicable to healthcare, financial services, government, and other regulated industries
• Reference guide — not a substitute for legal or compliance review

Prompt

Act as a compliance analyst. Provide a reference briefing on Copilot Chat data handling, web search behavior, and HIPAA/regulated-industry considerations. Produce the following:

---

### What Crosses the Tenant Boundary

Explain what data leaves the Microsoft 365 tenant when:
1. **Web search is enabled**: What queries are sent externally, what's cached, what's retained
2. **Web search is disabled**: What stays within the tenant, what still leaves
3. **File attachments are used**: How uploaded documents are processed and whether content exits the tenant
4. **Copilot in apps (Word, Outlook, Teams)**: Does data handling differ from standalone Copilot Chat?

For each scenario, specify:
- Data type (query text, file content, metadata)
- Destination (Microsoft cloud, third-party search providers)
- Retention period
- Whether it's used for model training

---

### BAA Coverage

Explain Microsoft's Business Associate Agreement (BAA) coverage for Copilot:
1. Which Copilot services are covered under the BAA
2. What constitutes Protected Health Information (PHI) in the Copilot context
3. Limitations and exclusions in the BAA
4. What customer responsibilities remain even with a BAA in place

---

### Regulated Industry Guidance

Provide compliance considerations for:
1. **Healthcare (HIPAA)**: PHI handling, audit trail requirements, patient data risks
2. **Financial Services (SOX, GLBA)**: Customer data, trading information, audit requirements
3. **Government (FedRAMP, ITAR)**: Classification boundaries, export controls, authorized environments
4. **EU (GDPR)**: Data residency, right to erasure, processing basis

For each: what's safe to use Copilot with, what requires additional controls, what should be avoided.

---

### Safe Usage Checklist

Produce a practical checklist organizations should follow:
1. **Before enabling Copilot**: Governance steps, BAA execution, policy updates
2. **Daily usage rules**: What employees can and cannot paste into Copilot
3. **Web search settings**: When to enable, when to disable, admin controls
4. **Monitoring & audit**: How to log and review Copilot usage for compliance
5. **Incident response**: What to do if sensitive data is accidentally submitted

---

### Risk Matrix

| Data Type | Safe with Web Search On | Safe with Web Search Off | Requires Additional Controls | Never Submit |
|---|---|---|---|---|

Populate for: PHI, PII, financial data, trade secrets, classified information, internal strategy, customer contracts, employee records.

---

End with a disclaimer: This is informational guidance, not legal advice. Consult your compliance team and Microsoft's current documentation before making policy decisions.

Usage Tips

• Compliance review: Share this output with your legal/compliance team before organizational rollout.
• Policy drafting: Use the "Safe Usage Checklist" section as a foundation for your internal Copilot acceptable-use policy.
• Admin settings: Work with your Microsoft 365 admin to configure web search toggles, data loss prevention (DLP) rules, and audit logging.
• Stay current: Microsoft updates Copilot data handling policies regularly — re-run this prompt quarterly to refresh guidance.
• Industry-specific: Add "Focus on healthcare/HIPAA" or "Focus on financial services/SOX" to narrow the analysis to your sector.
• Microsoft documentation: Cross-reference output with Microsoft's official Copilot data privacy documentation and current BAA terms.

Key References

• Microsoft Copilot data privacy and security documentation
• Microsoft Business Associate Agreement (BAA) for Microsoft 365
• HIPAA Security Rule §164.308 (administrative safeguards)
• GDPR Article 28 (processor agreements)
• NIST AI Risk Management Framework (for government use)

What This Is

A personal operations agent that protects your time, drafts in your voice, and preps briefings before you need them. Includes Word templates for fast setup.

The highest-ROI agent you can build. Set it up once, and it runs your daily prep automatically.

Setup Instructions

1. Open Copilot Studio
2. Create a new agent → Name it Chief of Staff
3. Paste the system prompt below
4. Upload your voice sample (3-5 recent emails or docs you're proud of)
5. Add the personalization block with your schedule and priorities
6. Connect Outlook, Calendar, and SharePoint data sources

System Prompt

You are the Chief of Staff — a personal operations agent that works at executive altitude.

Your role:
- Prep daily briefings: what needs attention, what can wait, what's at risk
- Draft communications in the leader's established voice and tone
- Pre-read documents and meetings, surfacing only what matters
- Protect the leader's time by filtering noise and flagging signal
- Anticipate needs based on calendar, recent activity, and priorities

Guidelines:
- Lead with decisions, not data dumps
- Use the leader's communication style — match tone, formality, and directness
- Always separate urgent from important
- Provide recommended actions, not just information
- When in doubt, err on the side of brevity
- Never fabricate context — flag gaps explicitly

Personalization Block

My profile:
- Role: [your title]
- Organization: [your org]
- Current top priorities: [2-3 items]
- Communication style: [direct/formal/casual — describe how you write]
- Meeting cadence: [key recurring meetings]
- People I report to: [names/roles]
- People who report to me: [names/roles]
- Things I delegate: [what you don't want to see]
- Things I always review: [what must come to you]

Voice Sample Setup

Upload 3-5 recent communications you're proud of (emails, docs, messages). The agent will learn your style from these. Recommended mix:

• One external communication (client/customer-facing)
• One internal leadership communication
• One quick reply (shows your casual tone)
• One formal document or briefing

Sample Prompts

Daily briefing

Prepare my morning briefing. What needs my attention today, what can wait, and what's at risk this week?

Draft a reply

Draft a reply to this thread. Agree to the timeline but push for the budget in writing. Match my tone.

Meeting prep

Prep me for my 2pm with [name]. What's the context, what was decided last time, and what should I push for today?

Document pre-read

Pre-read this document. Give me the bottom line, the risks, and the three questions I should ask in the review meeting.

Tips

• Schedule a daily morning briefing prompt to run automatically
• Update priorities monthly — stale context degrades output quality
• Use the "draft a reply" prompt for high-stakes emails before you send
• Build a weekly review prompt that runs Friday afternoon

What This Is

A virtual senior advisor for strategy thinking, career navigation, and blind-spot analysis. The Executive Coach helps you clarify your impact, see the bigger picture, and make moves with confidence.

Build this once in Copilot Studio. Use it whenever you need a sounding board that knows your context.

Setup Instructions

1. Open Copilot Studio
2. Create a new agent → Name it Executive Coach
3. Paste the system prompt below into the agent's instructions
4. Add your background/context using the personalization block
5. Test with the sample prompts at the bottom

System Prompt

You are the Executive Coach — a senior advisor who helps leaders think strategically, navigate career moves, and surface blind spots.

Your role:
- Act as a trusted sounding board for strategic decisions
- Help clarify impact and position it effectively
- Surface blind spots with direct, constructive feedback
- Frame career and organizational moves in terms of risk and upside
- Keep the tone professional, direct, and practical — no fluff

Guidelines:
- Use a consultant-style structure: categories, bullet points, actionable advice
- Match the user's tone: direct, helpful, professional
- When analyzing situations, separate facts from assumptions
- Always include at least one concrete next step
- Do not give generic advice — ground everything in the user's actual context

Personalization Block

Add this to the agent's instructions to ground it in your reality:

About me:
- Role: [your title]
- Organization: [your org]
- Key priorities right now: [list 2-3]
- Communication style: [how you prefer feedback]
- Context I want you to always keep in mind: [any recurring themes]

Sample Prompts

Strategy review

I'm considering [decision]. Walk me through the strategic implications, risks I might be missing, and what a smart move looks like here.

Career navigation

Based on my background and current trajectory, what's the highest-leverage career move I should make in the next 12 months? Be direct.

Blind-spot check

What are three things about my approach or communication that could hold me back at the next level? Keep it constructive but honest.

Positioning advice

I need to position [achievement/project] for maximum impact. How would you frame this for [audience]?

Tips

• Run this before important meetings or decisions as a pre-flight check
• Use it weekly for a structured self-review
• Save your best exchanges — they become a personal strategy log
• Update your personalization block quarterly as priorities shift

What This Is

An operational agent that tracks action items, pulls decisions from meetings, and gives strategic guidance to move work forward. Think of it as your personal program manager — always current, never forgets a commitment.

Build this once. Feed it meeting notes, email threads, or status updates. Get back clear ownership and next steps.

Setup Instructions

1. Open Copilot Studio
2. Create a new agent → Name it Portfolio Tracker
3. Paste the system prompt below into the agent's instructions
4. Connect your M365 data sources (Outlook, Teams, SharePoint)
5. Test with the sample prompts at the bottom

System Prompt

You are the Portfolio Tracker — an operational agent that tracks commitments, surfaces action items, and keeps work moving.

Your role:
- Extract action items from meeting notes, emails, and conversations
- Track ownership and deadlines for every commitment
- Surface overdue items and at-risk deliverables
- Provide strategic guidance on prioritization and sequencing
- Keep a running portfolio view of all active work

Guidelines:
- Always separate confirmed actions from open questions
- Assign clear owners and deadlines — flag when either is missing
- Use a structured format: Item | Owner | Deadline | Status | Risk
- Escalate blockers proactively with recommended next steps
- Keep the tone operational and direct — no fluff, just status

Personalization Block

About my work:
- Active projects: [list current projects]
- Key stakeholders: [names/roles]
- Reporting cadence: [weekly/biweekly/monthly]
- Escalation path: [who gets escalated items]
- Priority framework: [how you prioritize — e.g. RICE, MoSCoW]

Sample Prompts

Extract actions from a meeting

Here are the notes from today's standup. Extract every action item, assign owners where obvious, flag anything without a clear owner or deadline.

Portfolio status

Give me a current view of all open action items across my projects. Group by priority, flag anything overdue, and recommend what to tackle next.

Risk assessment

Looking at my current commitments, what's most at risk of slipping? What should I escalate or deprioritize?

Weekly digest

Summarize everything that moved forward this week, what's stalled, and what needs my attention in the next 48 hours.

Tips

• Run this after every meeting before you close the loop
• Set a weekly scheduled prompt for portfolio review
• Use it to prep for 1:1s — feed it your manager's questions and get a brief
• Connect it to your calendar so it knows your upcoming deadlines

What This Is

An agent that summarizes and links across your Microsoft 365 activity so you can see your real influence on projects, decisions, and people. Turns scattered activity into a clear impact narrative.

Use this before performance reviews, promotion cycles, or when you need to articulate your contribution.

Setup Instructions

1. Open Copilot Studio
2. Create a new agent → Name it Impact Auditor
3. Paste the system prompt below
4. Connect Outlook, Teams, SharePoint, and OneDrive data sources
5. Add the personalization block with your review cycle and key metrics

System Prompt

You are the Impact Auditor — an agent that maps activity to influence.

Your role:
- Scan across the user's M365 activity (emails, meetings, documents, chats)
- Identify patterns of influence: decisions shaped, problems solved, people mentored
- Quantify impact where possible: scope, frequency, outcomes
- Surface under-recognized contributions that deserve visibility
- Structure findings as a clear narrative suitable for reviews or self-assessments

Guidelines:
- Separate activity from impact — being busy is not the same as being influential
- Use evidence, not opinions: cite specific examples and outcomes
- Group findings by theme: strategic influence, operational impact, people development
- Always include a "visibility gap" section: contributions that happened but weren't seen
- Keep tone factual and constructive

Personalization Block

My context:
- Review cycle: [quarterly/annual/promotion]
- Key goals this period: [list 2-3]
- Metrics that matter in my org: [revenue, headcount, delivery, satisfaction, etc.]
- People I've mentored or influenced: [names/roles]
- Projects I've led or contributed to: [list]
- What I want to be recognized for: [specific outcomes]

Sample Prompts

Quarterly impact review

Review my activity over the last quarter. What were my top 5 contributions, and which ones flew under the radar?

Preparation for promotion

I'm preparing a promotion packet. Summarize my impact this year across strategy, delivery, and people. Include specific examples I can cite.

Influence mapping

Map my influence across the organization. Who do I regularly impact, what decisions have I shaped, and where are the gaps?

Visibility audit

What contributions have I made that key stakeholders probably don't know about? Suggest how to surface them.

Tips

• Run this 2 weeks before any review cycle
• Export the narrative and paste it into your self-assessment
• Use the "visibility gap" section to plan proactive communication
• Re-run quarterly to build a running impact log

What This Is

An executive presence coach that analyzes your communication patterns, identifies blind spots, and helps you project authority and confidence. Based on real communication science, not vibes.

Use this before high-stakes presentations, negotiations, or when you want to calibrate how you come across.

Setup Instructions

1. Open Copilot Studio
2. Create a new agent → Name it Presence Coach
3. Paste the system prompt below
4. Upload 3-5 samples of your communications (emails, docs, messages)
5. Add your target audience and context

System Prompt

You are the Presence Coach — an executive communication advisor grounded in presence science.

Your role:
- Analyze the user's communications for tone, clarity, and authority signals
- Identify patterns that undermine or strengthen executive presence
- Coach on calibration: reading the room, adjusting for audience
- Surface blind spots the user cannot see in their own writing
- Provide specific, actionable adjustments — not vague feedback

Guidelines:
- Frame feedback around three dimensions: gravitas, confidence, and connection
- Use concrete examples from the user's actual communications
- Separate observation from prescription: "here's what I see" then "here's what to try"
- Never attack the person — critique the communication, not the communicator
- Keep it direct but constructive: honest without being harsh

Personalization Block

My context:
- Role and level: [your title]
- Key audiences I communicate with: [executives, peers, direct reports, external]
- Upcoming high-stakes moments: [presentations, negotiations, reviews]
- My communication strengths: [what you already do well]
- My known weaknesses: [what others have told you]
- Cultural context: [org culture, regional norms]

Sample Prompts

Communication audit

Review these three emails I sent this week. What presence signals do they send? Where do I come across as strong, and where do I undermine myself?

Calibration check

I'm about to present to the board. Review this draft and tell me: does it project authority without arrogance? Where do I need to tighten up?

Blind-spot surfacing

Based on my communication samples, what are the top 3 blind spots I probably don't see in my own writing?

Audience adjustment

I need to deliver bad news to my team. Review this draft and help me adjust the tone: direct but not cold, honest but not alarming.

Tips

• Run this before any communication where perception matters
• Upload fresh samples monthly — your style evolves
• Use the "calibration check" before high-stakes sends
• Pair with the Redline Reviewer for a two-layer quality gate

What This Is

A document review agent that checks for tone, clarity, risk language, and strategic alignment before you send. Think of it as your pre-flight quality gate — catches what you'd miss in a tired review.

Run this on any high-stakes document, email, or deck before it leaves your hands.

Setup Instructions

1. Open Copilot Studio
2. Create a new agent → Name it Redline Reviewer
3. Paste the system prompt below
4. Add your organizational context and audience profiles
5. Test with a recent draft

System Prompt

You are the Redline Reviewer — a document quality gate for high-stakes communications.

Your role:
- Review documents for clarity, tone, and strategic alignment
- Flag risky language: overpromises, vague commitments, defensive phrasing
- Identify logical gaps, unsupported claims, and missing context
- Check audience fit: is this calibrated for who will read it?
- Suggest specific edits, not general feedback

Guidelines:
- Use a structured review: Clarity | Tone | Risk | Alignment | Recommendations
- Quote the exact text that needs changing, then provide the suggested replacement
- Separate "must fix" from "nice to improve"
- Flag anything that could be misread or cause unintended reactions
- Keep the review concise — the user needs to act on this fast

Personalization Block

My context:
- Role: [your title]
- Typical audiences: [executives, clients, team, board]
- Org tone: [formal/casual/direct — describe your org's communication style]
- Red lines: [topics or language to always flag — e.g. pricing, legal, commitments]
- What "good" looks like: [describe your standard for a clean send]

Sample Prompts

Pre-send review

Review this email before I send it to the executive team. Flag any risky language, tone issues, or gaps. Suggest exact replacements.

Deck review

Review this presentation for the board. What's unclear, what's overpromised, and what's missing that they'll ask about?

Risk scan

Scan this document for anything that could be misread or cause problems. Be specific about the exact text and why it's risky.

Audience calibration

This is going to [audience]. Is the tone and level of detail right? What should I add, remove, or reframe?

Tips

• Run this as the final step before any high-stakes send
• Pair with Presence Coach for a two-layer review: substance + perception
• Use the "risk scan" prompt for anything involving commitments or numbers
• Save your best redline notes — they become your personal style guide

What You Will Learn

This playbook equips you to operate Copilot Premium with executive precision. By the end, you will be able to:

• Assess data safety confidently — understand tenant boundaries, permission models, and audit capabilities
• Locate Copilot across every Microsoft 365 surface — chat, Office apps, Teams meetings, and mobile
• Achieve your first win within five minutes — catch up on context, clear inbox backlog, and walk into meetings prepared
• Apply the prompt formula — Goal + Context + Source + Format — to generate consistently high-quality outputs
• Treat Copilot as a chief of staff — personalize memory, configure agents, and schedule recurring prompts
• Navigate app-by-app capabilities — Outlook, Word, Excel, PowerPoint, and Teams with production-ready prompts
• Execute a 30-day adoption plan — structured weekly milestones from first login to sustained productivity gains

Why This Matters

Getting Time Back

Executives lose an average of 12–15 hours per week to context-switching, inbox management, and document synthesis. Copilot Premium is designed to reclaim that time by handling the preparatory work that precedes every decision.

The Chief of Staff Analogy

Think of Copilot not as a chatbot but as a chief of staff who:

• Reads everything you read, at your permission level
• Summarizes before you open the document
• Drafts responses before you compose
• Prepares briefing materials before you enter the room
• Remembers your preferences and working style

The difference between a tool and a staff member is continuity. Copilot maintains context across sessions, learns your communication patterns, and surfaces what matters before you ask.

Is My Data Safe?

Tenant Boundary

Copilot operates strictly within your organization's Microsoft 365 tenant. It does not cross tenant boundaries unless explicitly configured through shared channels or guest access provisions that already exist in your environment.

No Public Model Training

Your prompts, documents, emails, and conversations are not used to train public foundation models. Microsoft has committed to this boundary for all Copilot Enterprise and Premium deployments.

Respects Existing Permissions

Copilot inherits your existing Microsoft 365 permission model. It can only surface content you already have access to through SharePoint, OneDrive, Teams, or Exchange permissions. It does not escalate privileges.

Audit Trail

All Copilot interactions are logged in the Microsoft 365 audit log. Administrators can review:

• Which users accessed Copilot
• What content was referenced
• When prompts were executed
• Agent activities and outcomes

Where to Find Copilot

Copilot Chat (Primary Surface)

The central Copilot chat experience is available at ↗ Reference: Copilot chat portal. This is where you run cross-app queries, manage memory, and orchestrate multi-step workflows.

Office Applications

Teams Meetings

Copilot joins meetings where you are present (subject to admin policy). It captures transcripts, identifies decisions, and extracts action items in real time.

Mobile

The Copilot mobile app provides chat access on iOS and Android. Office app integrations vary by platform version.

Your First 5-Minute Win

Catch Up Fast

"Summarize the last 10 emails in my inbox that require my attention. Group by urgency and include a one-line recommended action for each."

Clear Your Inbox

"Draft responses to all emails in my inbox tagged 'Follow up' or 'Action Required'. Use a professional tone and keep each response under 4 sentences."

Walk In Ready

"I have a meeting with [name/team] at [time] about [topic]. Summarize any recent emails, documents, or chat messages related to this topic so I'm prepared."

These three prompts alone typically save 30–60 minutes of daily context-switching overhead.

The Prompt Formula

Consistent results come from structured prompts. Use this four-component formula:

Goal + Context + Source + Format

Full Example

"Draft a one-page executive summary (Goal) for a board meeting next Tuesday where the audience is non-technical (Context), based on the Q3 earnings report in my OneDrive (Source). Use bullet points, maximum 300 words, and include key metrics (Format)."

Copilot as Your Chief of Staff

Personalization

Tell Copilot how you work:

"When I ask for email drafts, keep them under 5 sentences and use a direct, no-fluff tone."

Memory

Copilot remembers instructions you give it across sessions. Use memory to store:

• Preferred communication style
• Recurring meeting topics
• Stakeholder context
• Decision-making frameworks

Agents

Copilot agents are persistent workflows that run on your behalf. Configure agents to:

• Monitor specific email threads daily
• Generate weekly status reports
• Track project milestones
• Surface relevant documents before meetings

Scheduled Prompts

Set recurring prompts that execute automatically:

"Every Monday at 8 AM, summarize my calendar for the week and flag any preparation needed."

Copilot by Application

Outlook

Word

Excel

PowerPoint

Teams

Perspectives and Insights

Contextual Filter

Ask Copilot to reframe content through a specific lens:

"Review this proposal through the lens of a skeptical CFO. What concerns would they raise?"

Devil's Advocate

Surface blind spots before decisions:

"Play devil's advocate on this strategy. What are the three strongest counterarguments?"

Voice Distiller

Match your communication style:

"Rewrite this in my voice. I prefer direct language, short paragraphs, and data-backed claims."

Your First 30 Days

Week 1: Foundation

• Complete first login and security setup
• Run the three 5-minute win prompts
• Configure Copilot memory with your working preferences
• Identify one recurring task to automate

Week 2: Application Depth

• Use Copilot in at least 3 Office applications daily
• Create your first scheduled prompt
• Run a meeting recap in Teams
• Draft one document entirely with Copilot assistance

Week 3: Advanced Patterns

• Apply the prompt formula to 5+ different tasks
• Configure a Copilot agent for recurring work
• Use perspectives and insights features
• Share one Copilot workflow with a peer

Week 4: Sustained Integration

• Audit time saved vs. Week 1 baseline
• Refine prompts based on what worked
• Document your top 5 go-to prompts
• Plan next-month adoption goals

Copilot Is Not Autopilot

What Copilot Does Well

• Summarizing large volumes of content
• Drafting first versions of documents and emails
• Identifying patterns in data
• Structuring information from multiple sources
• Translating technical content for different audiences

What Copilot Does Not Do

• Make decisions on your behalf
• Guarantee factual accuracy — always verify critical information
• Replace human judgment on sensitive matters
• Access content outside your permission boundary
• Understand organizational politics or unspoken context

The Human-in-the-Loop Principle

Every Copilot output should pass through three filters before action:

1. Accuracy check — Are the facts correct?
2. Tone check — Does this match how I would communicate?
3. Completeness check — Is anything missing that a human would catch?

If You Work Through an Assistant

Your executive assistant can operate Copilot on your behalf with proper licensing and permissions:

• Drafting: Your EA can use Copilot to prepare drafts for your review
• Meeting prep: Copilot can generate briefing materials before you enter the room
• Inbox management: Copilot can triage and categorize emails for your attention
• Calendar optimization: Copilot can identify scheduling conflicts and suggest adjustments

Best Practice

Establish a shared prompt library between you and your EA. This ensures consistency in tone, format, and decision-making frameworks across all Copilot-assisted outputs.

Resource Library

Session 1: Admin Center Foundations

Phased Rollout

Governance begins with controlled access. The Copilot Admin Center supports phased deployment across user cohorts:

Copilot Control System

The Copilot Control System is the centralized policy engine for managing AI capabilities across Microsoft 365. Key controls include:

• Chat scope — Define which data sources Copilot can access (chat, files, email, calendar)
• Content filters — Block specific document types or sensitivity labels from AI processing
• Feature toggles — Enable/disable specific Copilot capabilities per user group
• Network restrictions — Limit Copilot access to approved networks or locations

Guardrails

Guardrails are pre-configured safety boundaries that cannot be overridden by end users:

• Data residency enforcement — Ensure processing stays within required geographic boundaries
• Export controls — Prevent AI-generated content from leaving the tenant
• Prompt logging — Capture all prompts for audit and compliance review
• Usage monitoring — Track consumption patterns and flag anomalies

↗ Reference: Microsoft Learn: Copilot Admin Center documentation

Session 2: Copilot Studio + Power Platform Governance

Managed Environments

Managed Environments in Power Platform provide the foundation for Copilot Studio governance:

Data Loss Prevention (DLP)

DLP policies govern how Copilot Studio agents interact with connectors and data sources:

• Connector groups — Classify connectors as "business" or "non-business" and restrict cross-group data flow
• Custom connectors — Review and approve custom connectors before deployment
• Sensitive data scanning — Automatically detect and block PII, financial data, or regulated content
• Policy inheritance — Child environments inherit parent DLP policies unless explicitly overridden

Cost Model

Copilot Studio and Power Platform licensing follows a consumption-based model:

↗ Reference: Microsoft Learn: Power Platform governance documentation

Session 3: Purview for M365 + Agents

Sensitivity Labels

Microsoft Purview sensitivity labels classify and protect content across the tenant:

Data Loss Prevention (DLP) Policies

Purview DLP policies extend beyond Power Platform into Microsoft 365 content:

• Built-in policies — Pre-configured rules for credit cards, SSNs, health data, and financial information
• Custom policies — Define organization-specific sensitive data patterns
• Agent enforcement — Apply DLP rules to Copilot agent inputs and outputs
• Incident response — Automated workflows for DLP policy violations

Insider Risk Management

Insider risk management detects and responds to data misuse:

• User activity monitoring — Track unusual file access, downloads, or sharing patterns
• Case management — Investigate flagged activities with full audit trails
• Automated responses — Trigger alerts, block access, or initiate workflows based on risk scores
• Agent visibility — Monitor Copilot agent behavior for policy violations

Audit Logging

The Microsoft 365 audit log captures all Copilot-related activities:

• Prompt logging — Record every user prompt and Copilot response
• Content access — Log which documents Copilot referenced during processing
• Agent execution — Track agent triggers, inputs, and outputs
• Policy enforcement — Record when guardrails blocked or modified content

↗ Reference: Microsoft Learn: Microsoft Purview documentation

Session 4: SharePoint Advanced Management

Oversharing Detection

SharePoint Advanced Management (SAM) identifies content shared beyond intended boundaries:

• Access analytics — Visualize who has access to what, across SharePoint and OneDrive
• External sharing alerts — Flag documents shared with external users or guest accounts
• Permission inheritance breaks — Identify files with broken permission chains
• Copilot visibility mapping — Show what content Copilot can access for each user

Content Lifecycle Management

SAM enforces content retention and disposition policies:

Access Reviews

Regular access reviews ensure permissions remain appropriate:

• Scheduled reviews — Automate quarterly or annual access certification
• Just-in-time access — Grant temporary elevated permissions with automatic expiration
• Role-based reviews — Review access by job function rather than individual users
• Copilot impact — Access reviews directly affect what Copilot can surface to each user

Retention, Archiving, and Content Disposition (RAC/RCD)

RAC/RCD policies govern the lifecycle of content:

• Retention labels — Keep content for required periods (regulatory compliance)
• Legal holds — Preserve content during litigation or investigations
• Disposition approvals — Require human approval before deleting labeled content
• Copilot data minimization — Reducing stale content reduces Copilot's processing surface and improves relevance

↗ Reference: Microsoft Learn: SharePoint Advanced Management documentation

The Governance Stack

Quick Start for Executives

Weeks 1–2: Foundation

• Identify governance owner and core team (IT, security, legal)
• Audit current Microsoft 365 permission model
• Define sensitivity label taxonomy for your organization
• Configure Copilot Control System with default guardrails

Weeks 3–4: Policy Design

• Draft DLP policies for sensitive data categories
• Design managed environment strategy for Copilot Studio
• Establish access review cadence (quarterly minimum)
• Define cost tracking and budget allocation per department

Weeks 5–6: Pilot Governance

• Deploy Copilot to pilot group with full governance stack
• Monitor audit logs for policy violations and anomalies
• Conduct first access review for pilot participants
• Document governance lessons learned and adjust policies

Ongoing: Continuous Improvement

• Monthly governance team reviews of audit data
• Quarterly access reviews across all Copilot users
• Semi-annual policy updates based on new capabilities
• Annual governance framework assessment against regulatory changes

↗ Reference: Microsoft Learn: Copilot governance best practices

Executive Summary

Microsoft Agent 365 is the control plane for managing AI agents across your Microsoft 365 tenant. It is not a builder tool, not a runtime environment, and not a per-agent license. It is a per-user, tenant-wide governance layer priced at $15/user/month.

What Agent 365 Does

• Provides a centralized console for agent lifecycle management
• Enforces identity, authorization, and data access policies for all agents
• Logs agent activities for audit and compliance
• Scales with your user count, not your agent count

What Agent 365 Does Not Do

• Build agents (that is Copilot Studio)
• Execute agent workflows (that is the runtime plane: Power Automate, Logic Apps, Azure Functions)
• Replace Copilot Premium or Copilot for Microsoft 365 licenses

The Bottom Line

If you deploy AI agents in Microsoft 365, you need Agent 365 for governance. The cost is predictable: $15 × licensed users. Adding 100 agents does not increase the Agent 365 bill.

Three-Plane Mental Model

Understanding Agent 365 requires separating three distinct planes of operation:

Why the Separation Matters

Product-by-Product Relationship

CCS vs. Agent 365: The Critical Distinction

Copilot for Microsoft 365 (CCS)

• What it is: The AI assistant embedded in Word, Excel, PowerPoint, Outlook, Teams, and Loop
• What it does: Summarizes, drafts, analyzes, and generates content within M365 apps
• Licensing: Per-user annual subscription (~$30/user/month)
• Governance: Managed through Copilot Control System in Microsoft 365 admin center

Microsoft Agent 365

• What it is: The control plane for AI agent governance
• What it does: Manages agent lifecycle, enforces policies, provides audit trails
• Licensing: Per-user monthly subscription ($15/user/month)
• Governance: This IS the governance layer

Key Difference

CCS makes your employees more productive with AI. Agent 365 makes your AI agent deployment governable. You can have CCS without Agent 365 (if you have no custom agents). You should not deploy custom agents without Agent 365.

Architecture Overview

Agent Lifecycle

Identity

• Agents operate under managed identities or service principals, not individual user accounts
• Agent 365 tracks which human owns each agent and which humans can interact with it
• Identity federation supports Azure AD, Entra ID, and hybrid environments

Authorization

• Role-based access control (RBAC) governs who can create, modify, or delete agents
• Data access policies define what M365 content each agent can read or write
• Just-in-time authorization supports temporary elevated permissions with automatic expiration

Data Access

• Agents inherit the permission model of their owner or assigned service account
• Agent 365 enforces minimum privilege — agents access only what their policy allows
• Cross-tenant agent data access is controlled through B2B collaboration settings

Logging

• Every agent execution is logged with timestamp, trigger, input summary, and output summary
• Logs are retained per organizational retention policy (configurable)
• Audit logs integrate with Microsoft 365 compliance center and third-party SIEM tools

Licensing Model

Per-User, Tenant-Wide

Agent 365 is licensed per user, not per agent. Once a user has an Agent 365 license:

• They can create unlimited agents
• They can interact with any agent in the tenant
• The governance layer covers all agents regardless of count

Runtime Consumption Is Separate

Cost Example

The Agent 365 line item is flat. Runtime costs scale with usage.

Mixed Licensing Scenarios

Scenario 1: Copilot Premium + Agent 365

Scenario 2: Agent 365 Without Copilot

Scenario 3: Departmental Rollout

What Changes and What Doesn't

What Changes When You Add Agent 365

What Doesn't Change

GA vs. Preview Status

Generally Available (GA)

Preview (Subject to Change)

Production Guidance

• Build your governance foundation on GA features
• Test preview features in managed environments before production use
• Budget for GA features; treat preview capabilities as bonus value
• Review Microsoft's servicing policy for preview-to-GA transition timelines

Key Explanations by Audience

For Executives

"Agent 365 is the governance layer for AI agents. It costs $15/user/month regardless of how many agents you deploy. It ensures every agent is registered, logged, and compliant. Think of it as the audit function for your AI program — not the AI itself."

For IT Directors

"Agent 365 gives you centralized control over agent lifecycle, identity, authorization, and data access. It integrates with existing Microsoft 365 admin tools and Microsoft Purview. Licensing is per-user, tenant-wide — no per-agent fees. Runtime costs (Power Automate, Azure) are billed separately."

For Architects

"Agent 365 operates on the control plane, separate from build (Copilot Studio) and runtime (Power Automate, Azure Functions). It enforces RBAC on agent creation, defines data access scopes per agent, logs all executions, and supports automated access reviews. The architecture follows a three-plane model: build, runtime, control. Agent 365 is the control plane."

↗ Reference: Microsoft Learn: Agent 365 documentation ↗ Reference: Microsoft Learn: Copilot licensing guide ↗ Reference: Microsoft Learn: Power Platform licensing

Why Sizing by Seat Count Is Wrong Math

The most common Cowork deployment mistake is treating it like a software license: count seats, assign licenses, measure adoption. This approach fails because Cowork is not a tool — it is a work multiplier.

The Seat-Count Fallacy

The Cost of Over-Licensing

Deploying Cowork to users whose work doesn't match its capabilities creates three problems:

1. Wasted spend — Licenses sitting idle or used for trivial tasks
2. Credibility erosion — Users conclude "AI isn't useful" because their work shape doesn't benefit
3. Measurement noise — Adoption metrics look bad because the denominator includes mismatched users

The Right Metric

Instead of "how many seats," ask: "How many multi-step, multi-source, structured-deliverable workflows exist in this organization?"

Each answer to that question maps to one Cowork user. Everything else maps to Copilot Premium or no AI license at all.

The One Question That Separates Premium from Cowork

Ask every potential user this single question:

"Does your daily work require combining information from multiple sources, executing multi-step processes, and producing a structured deliverable?"

If Yes → Cowork Candidate

Examples:

• Analyst pulling data from three systems, building a model, and delivering a briefing
• Program manager tracking 12 workstreams, synthesizing status, and producing a steering committee deck
• Legal counsel reviewing contracts across departments, identifying risk patterns, and drafting policy updates

If No → Copilot Premium or None

Examples:

• User who primarily reads and replies to email (Premium handles this)
• User who creates simple documents from scratch (Premium handles this)
• User who attends meetings and takes notes (Premium handles this)
• Frontline worker executing defined tasks (no AI license needed)

The Decision Matrix

Cowork Personas

These are the roles that consistently generate Cowork-grade work:

Common Thread

Every Cowork persona shares three characteristics:

1. Input complexity — Information comes from 3+ distinct sources
2. Process complexity — Work involves 3+ sequential or parallel steps
3. Output structure — Deliverable follows a defined format (report, briefing, model, deck)

Who Does Not Need Cowork

Frontline Workers

Task Workers

Light Office Users

The Rule of Thumb

If the work can be described in a single sentence ("I write reports," "I manage calendars," "I answer phones"), Cowork is overkill. If it requires a paragraph ("I pull data from three systems, reconcile discrepancies, build a forecast model, and deliver a variance briefing to the CFO"), Cowork is the right fit.

Selection Method

Step 1: Tier Your Population

Segment your workforce into three tiers based on work shape:

Step 2: Use Signals, Not Nominations

Don't ask people to volunteer. Use objective signals to identify Tier 1 candidates:

Step 3: Run a Cohort

Deploy Cowork to a measured cohort (50–200 users) for 60–90 days:

• Baseline: Measure current productivity metrics before deployment
• Instrument: Track Cowork usage patterns, prompt complexity, and output quality
• Validate: Confirm that Cowork users are actually producing multi-step deliverables
• Adjust: Refine selection criteria based on cohort data

Step 4: Set a Payback Threshold

Define the minimum value threshold before expanding:

Step 5: Expand by Work Shape, Not Department

When scaling, add users who match the validated work shape profile — not by department or title:

Expansion Checklist

• Cohort data validates work shape hypothesis
• Payback threshold is met or exceeded
• Selection criteria are documented and measurable
• Governance stack (Agent 365, Purview) is ready for expanded user base
• Training materials are updated based on cohort learnings
• Budget is approved for next expansion wave

Summary

Cowork is a precision tool for a specific work shape. Sizing by headcount wastes money and erodes credibility. Sizing by work shape delivers measurable ROI and builds organizational trust in AI capabilities.

The formula: Identify Tier 1 users through objective signals → validate with a cohort → expand by work shape profile → measure against payback threshold.

How to Follow Along

Prerequisites

Enabling Agent Mode

1. Open Copilot chat in Excel or the standalone Copilot app
2. Navigate to Settings → Agent Mode
3. Toggle Agent Mode to ON
4. Confirm data access permissions (files, calendar, email as needed)
5. Agent Mode is now active — Copilot will execute multi-step workflows autonomously

Downloading Mock Data

The sample workbook contains five tabs representing a realistic operational dataset:

• Overtime_Raw — Weekly overtime hours by employee, department, and reason code
• Employee_Roster — Current headcount, role, department, and hire date
• Call_Volume — Daily inbound call volume by queue and priority level
• Staffing_Plan — Planned vs. actual FTE by week and department
• Budget — Monthly labor budget, actual spend, and variance by department

Data Description

Overtime_Raw

Employee_Roster

Call_Volume

Staffing_Plan

Budget

Analyst Prompt Pack

These six prompts walk through a complete analytical workflow in Agent Mode. Run them sequentially for the best experience.

Prompt 1: Prepare the Data Model

"Review all five tabs in this workbook. Identify relationships between tables (Employee_ID links Overtime_Raw to Employee_Roster; Department links across all tables; Date/Week_Ending links time dimensions). Create a summary table on a new tab called 'Data_Model' that documents each table, its key columns, and the relationships between them."

What this does: Establishes the data model foundation so subsequent prompts have shared context.

Prompt 2: Build the KPI Strip

"On a new tab called 'KPI_Dashboard', create a one-row KPI strip showing: (1) total overtime hours this quarter, (2) average overtime per employee, (3) total call volume this month, (4) average abandon rate, (5) current staffing gap (Planned_FTE minus Actual_FTE across all departments), and (6) total labor variance percentage. Pull live from the data, not hard-coded."

What this does: Creates a live dashboard header with six executive-level metrics.

Prompt 3: Generate Dashboard Visuals

"Below the KPI strip on the 'KPI_Dashboard' tab, create four charts: (1) overtime hours by department as a bar chart, (2) call volume trend over the last 30 days as a line chart, (3) staffing gap by department as a stacked bar showing planned vs. actual, and (4) budget variance by department as a waterfall chart. Size them to fit in a 2x2 grid."

What this does: Builds a visual dashboard with four complementary charts.

Prompt 4: Find the Story

"Analyze all five data tabs together. What is the single most significant finding when you connect overtime patterns, staffing gaps, call volume, and budget variance? Give me a three-paragraph narrative: (1) what is happening, (2) why it matters, and (3) what the data suggests is driving it."

What this does: Forces Copilot to synthesize across all data sources rather than analyzing one tab in isolation.

Prompt 5: SBAR for the CFO

"Format the finding from the previous prompt as an SBAR (Situation, Background, Assessment, Recommendation) memo addressed to the CFO. Keep it under 400 words. Use a formal executive tone. Include specific numbers from the data to support each section."

What this does: Translates analytical findings into a structured executive communication format.

Prompt 6: VP Talking Points

"Convert the SBAR memo into five bullet-point talking points suitable for a VP to use in a 15-minute budget review meeting. Each bullet should be one sentence, lead with the key fact, and end with the implication."

What this does: Adapts the same analysis for verbal delivery in a meeting context.

Executive Prompt Pack

These five prompts are designed for executives who want insights without building dashboards. Run them independently.

Prompt A: CFO Headline

"Look at the Budget and Overtime_Raw tabs. What is the single biggest cost risk in this quarter? Give me one sentence headline, three supporting data points, and one recommended action."

Expected output: A crisp risk statement with evidence and a clear next step.

Prompt B: Risk Identification

"Cross-reference Overtime_Raw with Staffing_Plan and Call_Volume. Identify the top three departments where overtime is high but staffing gaps are also high — meaning we are paying for overtime while still understaffed. List them with numbers."

Expected output: A prioritized list of departments with conflicting signals (high OT + high gap).

Prompt C: Month-over-Month Trend

"Compare this month's call volume, average handle time, and abandon rate against last month's numbers. Summarize the change as improving, stable, or degrading for each metric. Then give me an overall assessment in one sentence."

Expected output: A three-metric trend analysis with a net assessment.

Prompt D: Decision Options

"Based on the data across all tabs, present three options for reducing labor variance next quarter: (1) a staffing adjustment option, (2) a process improvement option, and (3) a budget reallocation option. For each, estimate the potential savings and the implementation risk level (high/medium/low)."

Expected output: Three structured options with estimated impact and risk.

Prompt E: Board Speaking Points

"Draft four speaking points for a board update on operational efficiency. Each point should be: one headline sentence, one supporting metric from the data, and one forward-looking statement. Keep the total under 200 words."

Expected output: Board-ready talking points with data backing and forward momentum.

Why This Demo Lands

For Executives

This walkthrough demonstrates that Copilot in Agent Mode can move from raw data to board-ready output in a single session. No Power BI. No pre-built models. No analyst handoff. The executive asks the question and receives a structured answer.

For IT Leaders

The demo shows Agent Mode executing multi-step workflows autonomously: reviewing data, identifying relationships, building charts, synthesizing findings, and formatting outputs. This is the difference between a chatbot and an agent — the latter does work, not just answers questions.

For Change Management

The Analyst Prompt Pack provides a repeatable training path. New users can follow the six prompts sequentially and see tangible output at each step. This reduces the "blank page" problem that kills AI adoption.

Key Takeaways

Next Steps

• Replace mock data with your actual operational dataset
• Customize prompts to match your department's KPIs and reporting cadence
• Schedule weekly Agent Mode sessions to build muscle memory
• Document your top-performing prompts in the Copilot Command Center Prompt Library

Why AI Can't Find Files Named Poorly

Copilot, agents, and search tools rely on file names and metadata to discover, classify, and surface content. A file named doc_final_v2_updated.xlsx tells the AI nothing about what is inside. A file named Q3_2026_labor_variance_by_department.xlsx tells it everything.

The Discovery Chain

The Cost of Poor Naming

• Copilot can't surface it — When you ask "Show me the Q3 budget analysis," a file named analysis.xlsx won't appear
• Agents can't use it — Custom agents rely on metadata to decide which files to read
• Governance breaks — Retention policies, sensitivity labels, and access reviews depend on accurate classification
• Search quality degrades — Microsoft 365 search weights file names heavily; garbage names = garbage results

Two Levels of AI Discoverability

Level 1: The File Title

The file name is the primary signal. It should encode:

Pattern: [Period]_[Subject]_[Scope]_[Type].[ext]

Example: Q3_2026_labor_variance_by_department_report.xlsx

Level 2: The Top-of-File Metadata Block

For documents that serve as reference material, policy documents, or static content, add a structured metadata block at the top. This gives AI systems machine-readable context that file names alone cannot provide.

Good vs. Bad File Names

Bad Names

Good Names

Naming Convention Quick Reference

The Metadata Block

Place this block at the top of any static reference document, policy, guide, or template. Use YAML-style key-value pairs for machine readability.

Example Metadata Block

---
title: "Q3 2026 Labor Variance Analysis Guide"
doc_type: "analysis_guide"
owner: "Finance Operations Team"
topics: ["labor_cost", "variance_analysis", "budgeting", "quarterly_review"]
effective_date: "2026-07-01"
version: "1.2"
status: "published"
review_date: "2026-10-01"
access_level: "internal"
related_docs:
  - "FY27_budget_forecast_by_department.xlsx"
  - "Q2_2026_labor_variance_summary.md"
---

Field Reference

Where to Use Metadata Blocks

You Don't Need This on Every File

Focus on Static Content

Apply rigorous naming and metadata to files that:

• Persist beyond a single use — Reference materials, policies, guides
• Are consumed by multiple people — Shared templates, departmental standards
• Feed into AI workflows — Data sources for agents, knowledge bases for RAG
• Have compliance requirements — Documents subject to retention, audit, or review

Skip It for Transient Content

Don't over-engineer naming for:

• Daily drafts — File name with date is sufficient
• Meeting notes — Date + meeting name in the filename
• Personal scratch work — If only you use it, your memory is enough
• Temporary exports — Delete after use; no metadata needed
• Attachments in email threads — Context lives in the conversation

The 80/20 Rule

20% of your files are reference-grade content that AI needs to find reliably. Spend your naming and metadata effort there. The remaining 80% are transient and don't need machine-readable structure.

Where This Is Headed: AI-First File Structures

The Trend

Organizations are moving toward AI-first file structures where every document is designed to be discovered, classified, and consumed by AI systems as well as humans. This means:

What to Prepare For

• Automatic classification — AI will read metadata blocks to apply sensitivity labels, retention policies, and access controls automatically
• Agent knowledge bases — Custom agents will index documents by metadata topics, not folder paths
• Cross-tenant discovery — B2B collaboration will rely on standardized metadata for safe content sharing
• Compliance automation — Regulatory requirements will be enforced through metadata fields, not manual processes

Start Now

• Audit your top 50 reference documents — rename and add metadata blocks
• Establish a naming convention standard for your team
• Train new hires on AI-ready file practices from day one
• Document your metadata taxonomy so it's consistent across departments

Summary

AI can only find what it can identify. File names are the first signal. Metadata blocks are the second. Invest in both for the 20% of files that matter most — your reference-grade, shared, compliance-sensitive content. The remaining 80% can stay simple.

The formula: Name files with [Period]_[Subject]_[Scope]_[Type]. Add metadata blocks to static reference documents. Skip the rest.