PROMPT LIBRARY.

You are preparing a board update for [YOUR NAME], [ROLE] at [COMPANY].

Context from last board meeting:
[PASTE PREVIOUS BOARD UPDATE OR KEY DECISIONS]

Current state:
[PASTE RAW NOTES, METRICS, OR BULLET POINTS FROM THE PAST MONTH]

Generate a board update with exactly four sections:

TODAY VS. PLAN
What actually happened vs. what was committed. Numbers first, narrative second. Flag any variance over 10%.

KEY DECISIONS MADE
What was decided, who decided it, and what the expected impact is. One bullet per decision. If a decision was reversed or delayed, say so.

RISKS & OPEN QUESTIONS
The three things that could go wrong in the next 90 days. For each: what it is, probability (high / medium / low), and what's being done about it. Don't soften.

NEXT PERIOD COMMITMENTS
Exactly what will be true at the next board meeting. Not goals — commitments. If you're not willing to be accountable for it, don't put it here.

Keep the whole update under 400 words. Write for a reader who has read the last three updates and wants only the delta.

I need to make a decision about [DESCRIBE THE DECISION IN ONE SENTENCE].

Background:
[DESCRIBE THE SITUATION — 3-5 sentences, including constraints and what you know]

My constraints: [BUDGET, TIMELINE, HEADCOUNT, OR OTHER HARD LIMITS]
My non-negotiables: [WHAT CANNOT CHANGE OR BE COMPROMISED]

Generate exactly three options. For each:

OPTION [N]: [Name in 3-5 words]
What it means: [One sentence — what actually happens if we choose this]
Upside: [The best case — specific, no hedging]
Risk: [The real downside — don't soften]
Decision trigger: [What would make this option clearly right]
Speed to value: [Fast / Medium / Slow]
Reversibility: [Easy / Hard / Irreversible]

Then:

RECOMMENDATION
Which option you'd choose and the actual reason why. One paragraph. Take a position — don't hedge or present "it depends."

Identify the highest-value AI use cases for [COMPANY NAME OR DESCRIPTION].

Company context:
- Industry: [INDUSTRY]
- Revenue: [REVENUE RANGE]
- Headcount: [HEADCOUNT]
- Core product or service: [WHAT THEY SELL]
- Main cost drivers: [TOP 2-3 COST LINES]
- Main revenue drivers: [TOP 2-3 REVENUE LINES]
- Current AI usage: [WHAT THEY ALREADY USE, OR "NONE"]

Generate the top 5 AI use cases, ranked by estimated P&L impact.

For each:

USE CASE [N]: [Name]
Where it sits: [Revenue growth / Cost reduction / Margin improvement]
What it does: [One sentence — what the AI actually does in practice]
P&L impact: [Estimated range — be specific, flag key assumptions]
Data needed: [What data is required and whether it likely exists]
Change complexity: [Low / Medium / High — and the main reason]
Kill risk: [What most likely prevents this from working]
Time to value: [Weeks / Months / Quarters]

At the end:

SEQUENCING RECOMMENDATION
Which use case to start with and why. Weight: data readiness, change complexity, speed to validate.

It's [DAY, DATE]. Help me reset priorities for the week.

My top 3 strategic priorities this quarter:
1. [PRIORITY 1]
2. [PRIORITY 2]
3. [PRIORITY 3]

This week's calendar (rough is fine):
[PASTE OR LIST YOUR MEETINGS]

Open items and loose ends from last week:
[LIST ANYTHING UNFINISHED OR PENDING]

New things that landed recently:
[LIST ANY NEW REQUESTS, ESCALATIONS, OR PROJECTS]

Generate:

THIS WEEK'S THREE THINGS
The three outcomes that matter most this week. Not tasks — outcomes. If these three things are true by Friday, the week was a success. Rank them.

CALENDAR AUDIT
For each meeting: Attend / Delegate / Cancel / Shorten. One sentence of reasoning for anything that isn't a clear Attend.

WHAT TO DROP THIS WEEK
What to actively not do. Every priority is also a de-prioritization — name what's being sacrificed.

NOT FOR ME
Anything on this list that should be delegated immediately, and to whom.

Audit my calendar for the week of [DATE RANGE].

My role: [YOUR ROLE]
My top priority this week: [ONE SENTENCE]

Meetings:
[PASTE YOUR CALENDAR — one meeting per line: Time · Title · Attendees if known]

For each meeting:
1. Is my presence actually required, or could someone else attend?
2. Could this be resolved async — email, written update, Loom?
3. If I must attend: what is the specific decision or input only I can provide?

Output for each meeting:
MEETING: [Title]
VERDICT: [Must Attend / Delegate / Cancel / Async]
REASON: [One sentence]
ACTION: [If delegating or cancelling — the exact message to send, ready to copy-paste]

At the end:
TIME RECLAIMED: Total hours freed if all recommendations are followed.
USE IT FOR: What to redirect that time toward, based on my top priority.

I just made a decision. Help me log it properly.

The decision: [WHAT WAS DECIDED]
When: [DATE]
Who was involved: [WHO HAD INPUT OR WAS AFFECTED]
What triggered it: [WHAT FORCED THE DECISION NOW]

Context I had at the time:
[WHAT YOU KNEW — data, opinions, constraints, what was uncertain]

What I chose NOT to do:
[THE ALTERNATIVES THAT WERE CONSIDERED AND REJECTED]

Generate a decision log entry:

DECISION: [One sentence, precise and specific]
DATE: [Date]
DECIDER: [Who made the final call]
TRIGGER: [What forced the decision]
OPTIONS CONSIDERED: [Bullet list with one-line rationale for each rejection]
RATIONALE: [Why this option. The actual reason — 2-3 sentences, not the PR version]
EXPECTED OUTCOME: [What you expect to be true in 90 days if this was right]
KILL CONDITION: [What would tell you the decision was wrong and should be reversed]
REVIEW DATE: [Specific date to check the outcome]

I'm considering building [FEATURE NAME].

One-sentence description: [WHAT IT DOES]
Who it's for: [TARGET USER]
The hypothesis: [WHY WE THINK THEY WANT IT]
Resources required: [ROUGH TEAM SIZE AND TIMELINE]

Before we commit to build, I need explicit kill criteria.

PRE-BUILD KILL CONDITIONS
What would stop us from starting? Name specific blockers — data gaps, market signals, resource conflicts.

ALPHA KILL CONDITIONS
What does failure look like in the first 5-20 design partners? Name specific behaviors, usage patterns, or feedback that means we stop. Avoid language like "low engagement" — define what low means in numbers.

SCALE KILL CONDITIONS
What does failure look like post-launch? Specific metrics, thresholds, and timeframes.

RESOURCE KILL CONDITION
At what point does team size or time invested make the opportunity cost unacceptable?

HARDEST KILL TO ENFORCE
Which condition are we most likely to rationalize away? Flag it — that's where discipline will be tested.

[COMPETITOR NAME] just [DESCRIBE WHAT THEY ANNOUNCED OR LAUNCHED].

Our context:
- Our product: [WHAT WE SELL]
- Our customers: [WHO BUYS FROM US AND WHY]
- Our main differentiator: [WHY CUSTOMERS CHOOSE US]
- Exposure: [ANY DEALS OR RENEWALS AT RISK IN THE NEXT 90 DAYS]

Analyze this move:

THREAT LEVEL: [High / Medium / Low] — one sentence on why.

CUSTOMER IMPACT
Which customer segments are most exposed? What conversation will sales and CS hear in the next 30 days?

WHAT THEY GOT RIGHT
Where does this address a real gap we have? Don't dismiss it — a competitor can be right.

RESPONSE OPTIONS
Three options, ranging from fast to structural:
1. [Immediate — days]: [What it is, what it costs, what it signals]
2. [Short-term — weeks]: [What it is, what it costs, what it signals]
3. [Strategic — months]: [What it is, what it costs, what it signals]

RECOMMENDED MOVE
Which option (or combination) and the actual reason. Take a position.

WHAT NOT TO DO
The reactive move that would be tempting but wrong.

I need to hire for [ROLE NAME].

Why this role exists now: [THE ACTUAL PROBLEM BEING SOLVED — not the job description version]
Reports to: [MANAGER]
Team today: [WHO'S ALREADY ON THE TEAM]
Budget range: [COMPENSATION RANGE]
Location: [IN OFFICE / HYBRID / REMOTE]

The three outcomes this person owns in year one:
1. [OUTCOME 1 — specific and measurable]
2. [OUTCOME 2]
3. [OUTCOME 3]

What failure looks like in this role: [BE SPECIFIC]
What we've tried before that didn't work: [IF APPLICABLE]

Generate:

THE ROLE IN ONE SENTENCE
What this person does and why it matters. For operators reading it, not HR processing it.

WHAT WE'RE ACTUALLY LOOKING FOR
5-7 bullets based on outcomes, not credentials. What does the right person's track record look like?

WHAT WE'RE NOT LOOKING FOR
3 bullets on the profile that interviews well but fails in this role. Name the anti-pattern.

THE INTERVIEW TEST
One real scenario that reveals whether someone can actually do this job. Not a brain teaser — a realistic work situation from this role.

THE PITCH
What would make the right candidate say yes. The actual reason to join — beyond comp.

Stress test this org structure.

My organization:
[DESCRIBE YOUR ORG — roles, reporting lines, team sizes. Rough is fine. A paragraph or a list works.]

Context:
- Stage: [STARTUP / GROWTH / MATURE / RESTRUCTURING]
- Top priority for the next 6 months: [ONE SENTENCE]
- Known tensions or issues: [ANYTHING YOU ALREADY KNOW IS SUBOPTIMAL]

Identify:

SINGLE POINTS OF FAILURE
Which roles or people, if they left tomorrow, would cause the most damage? What specifically breaks?

COVERAGE GAPS
What important work isn't clearly owned by anyone? Where does accountability get diffuse?

WRONG-LEVEL WORK
Where are senior people doing work that should sit lower? Where is junior work being escalated unnecessarily?

STRUCTURAL MISALIGNMENTS
Where does the org structure fight the strategy? Any reporting lines that create the wrong incentives?

REDUNDANCY
Any overlap in ownership or duplication of effort that's creating drag?

HIGHEST-IMPACT FIX
The one structural change with the most impact in the next 90 days. Name roles, not just patterns.

Help me prepare for a performance conversation.

Person: [NAME OR DESCRIPTION — e.g., "a senior PM, 2 years at the company"]
Situation: [WHAT'S BEEN HAPPENING — be honest. Good or bad.]
My goal: [WHAT I WANT TO BE TRUE AFTER THIS CONVERSATION]
What I think they'll say: [THEIR LIKELY FRAMING OR RESPONSE]
What I find difficult about this: [YOUR HONEST ANSWER]

Generate:

OPENING
The first sentence I should say. Not a buffer — the actual opening. Direct, sets the tone for an honest conversation.

THE CORE MESSAGE
What I need them to understand by the end. One paragraph. Write it as I'd say it, not as HR would write it.

THE HARD QUESTION
The question I'm probably avoiding but should ask. The one that opens the real conversation.

HOW TO HANDLE THEIR LIKELY RESPONSES
Three likely responses from them, and what I should say in return. Not scripts — anchors.

THE CLOSE
What I commit to, and what I'm asking them to commit to. Specific and time-bound.

WHAT NOT TO DO
The thing I'm most likely to do in the moment that will make this conversation less productive.