Claude Prompt Library

30 Claude Prompts That Build Board Dashboards

30 copy-paste prompts

Describe the pack you need and Claude returns a working board dashboard as one self-contained HTML file with sample numbers and charts you can click in the chat. Prompts for KPI packs, ARR, cash, OKRs, risks, and one-pagers. For a board meeting, not an ops floor.

In short: This page contains 30 copy-paste ready prompts, organized into 6 categories with a description and pro tip for each. The first 5 prompts are free instantly, no signup needed. Hand-curated and tested by the AI Academy team.

Louis Corneloup
By Louis Corneloup ยท Founder, Techpresso
Last updated ยทHand-curated & tested by the AI Academy team

KPI Pack

5 prompts

Eight Card Board Snapshot

1/30

โœจ What it does

Claude builds an eight-card board snapshot dated [BOARD DATE] with plan, prior, RAG, and sparklines as a working screen you can click in the chat. After you click a red KPI card, replace the sample numbers with this meeting's pack.

You are a CFO who builds one-screen board KPI packs a director can read without a decoder ring. <context> I need an eight-card board snapshot as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. This is for a board meeting, not an ops floor. </context> <inputs> - Company name: [COMPANY NAME] - Board date: [BOARD DATE] - Eight metrics with actual, plan, and prior: [E.G. ARR, NET NEW ARR, GROSS MARGIN, NET BURN, RUNWAY, CASH, NRR, HEADCOUNT] - RAG rule: [E.G. GREEN WITHIN 5% OF PLAN, AMBER WITHIN 15%, ELSE RED] - One sentence of narrative: [OR ASK CLAUDE TO DRAFT FROM THE NUMBERS] </inputs> <task> Build a single-screen pack dated [BOARD DATE] with: a header and a one-line narrative; a grid of eight KPI cards, each with name, actual, plan, versus plan %, versus prior, a red/amber/green status, and a sparkline; a count banner of how many metrics are green, amber, and red; and a click-to-expand footnote on each card that states the definition. Compute RAG from the rule constants. Clicking a card opens the footnote and dims the others. Fit the grid on a laptop screen without scrolling where you can. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - RAG is computed from actual versus plan against the threshold constants. - Status uses color plus the word green, amber, or red. - Responsive; accessible contrast; print-friendly light background. - Do not add warehouse, queue, or ticket widgets. Directors only. </constraints> <format> Return the full HTML file, then explain the RAG rule and which metric array to replace before the meeting. </format>

๐Ÿ’ก

Pro tip: Lock the eight metric names to last quarter's pack. Changing definitions mid-year is how a board stops trusting the page.

Plan Versus Actual Scorecard

2/30

โœจ What it does

Claude builds a plan versus actual scorecard from [ANNUAL PLAN] with implied full-year and a misses bridge as a working screen you can click in the chat. First click an amber plan line, then replace the sample numbers with the live plan.

You are an FP&A lead who builds annual-plan scorecards for a board, not a weekly flash for operators. <context> I need a plan versus actual scorecard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. Board altitude only. </context> <inputs> - Company name: [COMPANY NAME] - Annual plan lines: [ANNUAL PLAN] - Elapsed fraction of the year: [E.G. 0.58] - RAG rule: [E.G. GREEN WITHIN 5%, AMBER WITHIN 12%, ELSE RED] - Currency: [E.G. USD] </inputs> <task> Build the dashboard with: a KPI row (lines on plan, amber, red, full-year outlook versus plan); a table from [ANNUAL PLAN] with plan, actual, percent of year elapsed, implied full-year, variance to plan, and RAG; a bar of implied full-year versus plan; a bridge of the largest misses; and a callout of the one line that needs a decision this meeting. Implied full-year is actual divided by elapsed fraction when the line is a flow (revenue, spend), and actual when the line is a stock (cash, headcount). Clicking an amber or red row opens a one-line cause field already filled with sample text you can edit later. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Flow versus stock is a field on each line; do not annualize cash. - RAG computed from implied full-year versus plan. - Currency consistent; accessible contrast; print-friendly. </constraints> <format> Return the full HTML file, then explain the flow-versus-stock rule and which plan array to replace. </format>

๐Ÿ’ก

Pro tip: Mark cash and headcount as stock. Annualizing a bank balance is how a mid-year pack tells the board you will have twice the cash.

Quarter Versus Prior Quarter

3/30

โœจ What it does

Claude builds a quarter versus prior quarter pack for [COMPANY NAME] with deltas, mix notes, and a one-line story as a working screen you can click in the chat. Click the biggest swing, then replace the sample numbers with both quarters.

You are a CFO who builds quarter-versus-prior packs so a director can see what actually moved. <context> I need a quarter versus prior quarter dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. For a board meeting, not a weekly ops review. </context> <inputs> - Company name: [COMPANY NAME] - This quarter label: [E.G. Q2 2026] - Prior quarter label: [E.G. Q1 2026] - Eight to twelve lines with this-quarter and prior-quarter values: [E.G. ARR, NET NEW, GROSS MARGIN, BURN, CASH, NRR, HEADCOUNT, WIN RATE] - Currency: [E.G. USD] </inputs> <task> Build the dashboard with: a header for [COMPANY NAME] and both quarter labels; a KPI row of the four largest absolute deltas; a grouped bar of this quarter versus prior for every line; a variance table with $ and % and a favorable/unfavorable flag; a one-line story that names the single biggest swing; and a mix note if revenue or ARR mix is in the inputs. Compute every delta from labeled constants. Clicking a table row highlights that bar and rewrites the story line to that metric. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Deltas are computed, not hardcoded; biggest swing is computed. - Status uses color plus a word (up, down, flat). - Print-friendly; accessible contrast; no ticket queues or warehouse widgets. </constraints> <format> Return the full HTML file, then say which two arrays to replace and which swing the sample points to first. </format>

๐Ÿ’ก

Pro tip: Keep the same eight lines as last quarter. A new row that only exists this pack will look like you hid a miss.

RAG Traffic Light Pack

4/30

โœจ What it does

Claude builds a RAG traffic light pack from [METRIC LIST] using your green, amber, and red rule as a working screen you can click in the chat. Once you click a red metric, replace the sample numbers with the figures you will defend.

You are a board secretary who builds RAG packs that survive a skeptical director. <context> I need a RAG traffic light pack as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. Meeting room, not a floor screen. </context> <inputs> - Company name: [COMPANY NAME] - Metrics to score: [METRIC LIST] - Actual and plan (or target) per metric: [OR INVENT PLAUSIBLE ONES] - RAG rule: [RAG RULE] - Owner per metric: [NAMES OR ROLES] </inputs> <task> Build the dashboard with: a header and a count banner (green / amber / red); a grid of cards from [METRIC LIST], each with actual, plan, variance, RAG word, owner, and a one-line definition; a stacked bar of status counts; a red-first table sorted worst to best; and a click that opens the definition plus the rule that produced that color. Compute every status from [RAG RULE] constants. A filter chip shows only red, only amber, or all. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Color is never the only cue; print the word green, amber, or red. - Filters must hide cards; do not hardcode the banner counts. - Print-friendly light background; accessible contrast. </constraints> <format> Return the full HTML file, then restate the RAG rule in one line and say which array to replace. </format>

๐Ÿ’ก

Pro tip: Write the RAG rule in numbers, not vibes. If green is a feeling, two directors will argue the same card.

Board Narrative Header

5/30

โœจ What it does

Claude builds a board narrative header around [ONE LINE STORY] with eight supporting cards and footnotes as a working screen you can click in the chat. Click a footnote, then replace the sample numbers so the story still matches the math.

You are a CEO who writes the one line the board should remember, then proves it with eight numbers. <context> I need a board narrative header as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. This is the opening page of a board pack, not an ops wall. </context> <inputs> - Company name: [COMPANY NAME] - Board date: [BOARD DATE] - The one line I will stand behind: [ONE LINE STORY] - Eight supporting metrics with actual and plan: [METRIC LIST] - What I will not claim: [OFF-LIMITS] </inputs> <task> Build the dashboard with: a large header that prints [ONE LINE STORY] and [BOARD DATE]; eight supporting cards that must be consistent with that line; a small "proof" strip that lists which three cards carry the story; footnotes on each card; and a warning callout if a sample card would contradict the story (compute that, do not hide it). Clicking a card expands the footnote and highlights whether it supports, is neutral to, or tensions the header line. Keep the layout to one laptop screen. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Do not invent a second story. If a metric would break [ONE LINE STORY], flag it. - Honor [OFF-LIMITS]. No warehouse or queue widgets. - Print-friendly; accessible contrast. </constraints> <format> Return the full HTML file, then list which cards support the line and which constants to replace so the story still ties. </format>

๐Ÿ’ก

Pro tip: Write the story as a fact the numbers can prove. A vibe in the header and a miss in the cards will get you twenty minutes of distrust.

ARR and Growth

5 prompts

ARR Movement Bridge

6/30

โœจ What it does

Claude builds an ARR movement bridge from [ARR BRIDGE] with new, expansion, contraction, and churn as a working screen you can click in the chat. After you click a month on the waterfall, replace the sample numbers with your ARR movements.

You are a SaaS CFO who builds ARR bridges a director can audit in one sitting. <context> I need an ARR movement bridge as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. Board altitude: movements and mix, not a rep scoreboard. </context> <inputs> - Company name: [COMPANY NAME] - Starting ARR and ending ARR: [ARR BRIDGE] - Movement buckets: [E.G. NEW, EXPANSION, CONTRACTION, CHURN, REACTIVATION] - Period: [E.G. THIS QUARTER AND LAST FOUR QUARTERS] - Currency: [E.G. USD] </inputs> <task> Build the dashboard with: a KPI row (ending ARR, net new ARR, net new versus plan if given, quick ratio); a waterfall for the selected period that starts at beginning ARR, applies each bucket in [ARR BRIDGE], and lands on ending ARR; a four-quarter trend of ending ARR; a mix stacked bar of new versus expansion versus churn; and a table of the largest expansion and churn movements. A period toggle rebuilds the waterfall from that period's movements. Compute ARR as the constant you are given, not as MRR times 12 unless I only give MRR. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Each waterfall must reconcile beginning plus movements to ending; flag a break. - Churn and contraction render as negatives; cards recompute from the selected period. - Currency consistent; accessible contrast; print-friendly. </constraints> <format> Return the full HTML file, then explain the bridge math and which movement arrays to replace. </format>

๐Ÿ’ก

Pro tip: Give beginning and ending ARR that the movements can actually add to. A bridge that does not tie will die in the first question.

Net New ARR Versus Plan

7/30

โœจ What it does

Claude builds a net new ARR versus plan view from [ARR PLAN] with pace, gap, and a required monthly run rate as a working screen you can click in the chat. Click the lagging month, then replace the sample numbers with booked ARR.

You are an FP&A partner who builds net-new ARR versus plan for the board, not a weekly forecast huddle. <context> I need a net new ARR versus plan dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. Directors, not a sales floor. </context> <inputs> - Company name: [COMPANY NAME] - Period plan for net new ARR: [ARR PLAN] - Monthly net new actuals: [OR INVENT] - Months elapsed and months remaining: [E.G. 2 OF 3] - Currency: [E.G. USD] </inputs> <task> Build the dashboard with: a KPI row (plan, actual, gap, implied monthly pace versus required pace); a grouped bar of plan versus actual by month; a cumulative line of actual versus a straight-line plan; a remaining-gap callout; and a table of months with plan, actual, gap, and on/off-pace status. Compute required monthly pace from remaining gap and remaining months. Clicking a month focuses the table and the callout. Handle a closed period (0 months remaining) without dividing by zero. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Pace and gap are computed from plan, actual, and remaining months. - Off-pace months flagged with color and the word off-pace. - Currency consistent; accessible contrast; print-friendly. </constraints> <format> Return the full HTML file, then explain the required-pace formula and where to paste monthly actuals. </format>

๐Ÿ’ก

Pro tip: Enter months elapsed and months remaining yourself. A calendar that assumes a 30-day month will lie in a 91-day quarter.

NRR Board View

8/30

โœจ What it does

Claude builds an NRR board view from [SEGMENTS] with GRR, NRR, and a weakest-segment callout as a working screen you can click in the chat. First click the weakest segment, then replace the sample numbers with your retention file.

You are a SaaS CFO who builds NRR pages a board can trust without a seminar. <context> I need a net revenue retention board view as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. One page for directors, not a CS ops queue. </context> <inputs> - Company name: [COMPANY NAME] - Segments: [SEGMENTS] - Period: [E.G. LAST 12 MONTHS] - Beginning MRR or ARR per segment plus expansion, contraction, churn: [OR INVENT] - Target NRR: [E.G. 110%] </inputs> <task> Build the dashboard with: a KPI row (GRR %, NRR %, logo churn %, customers lost); a line of GRR and NRR over time with a 100% reference and the target; a segment table from [SEGMENTS] with customers, beginning revenue, GRR, NRR, logo churn; a bar of NRR by segment; and a callout of the weakest segment versus target. Compute GRR as (beginning - churned - contraction) / beginning and NRR as (beginning - churned - contraction + expansion) / beginning. NRR at or above 100% reads as healthy green. Clicking a segment focuses the table and the callout. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - GRR and NRR are computed from movement constants; show the formulas in comments. - Weakest segment is the lowest NRR, computed, not hardcoded. - Accessible contrast; print-friendly; 100% line labeled. </constraints> <format> Return the full HTML file, then explain GRR versus NRR and how to paste segment movements. </format>

๐Ÿ’ก

Pro tip: Keep logo churn and revenue churn on separate cards. Losing small accounts while NRR stays above 100% is a different problem than losing your top logo.

New Versus Expansion Bookings

9/30

โœจ What it does

Claude builds a new versus expansion bookings mix from [BOOKING MIX] with quota split and largest deals as a working screen you can click in the chat. Click the thinner type, then replace the sample numbers with this quarter's bookings.

You are a revenue CFO who builds new-versus-expansion bookings pages for the board. <context> I need a new versus expansion bookings dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. Board mix, not a rep activity board. </context> <inputs> - Company name: [COMPANY NAME] - Booking types and this-period amounts: [BOOKING MIX] - Quota or plan split if you have one: [E.G. 60% NEW, 40% EXPANSION] - Period: [E.G. THIS QUARTER] - Largest deals to name or anonymize: [OR INVENT PLAUSIBLE ONES] </inputs> <task> Build the dashboard with: a KPI row (total bookings, new, expansion, mix % versus the plan split); a stacked bar of bookings by type over the last four periods; a donut of this period; a table of the largest deals in each type; and a callout of which type is thin versus plan. Generate sample bookings with a type field in labeled JS arrays and compute the mix from that field. Clicking a type filters the deal table. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Mix percentages recompute from the bookings array. - Type colors stay consistent across every chart; currency formatted consistently. - No pipeline coverage or activity widgets. Bookings only. </constraints> <format> Return the full HTML file, then explain how the mix is computed and which array to replace. </format>

๐Ÿ’ก

Pro tip: Tag each booking the way the board already hears it (new, expansion, renewal). A third label invented for this page will restart the mix argument.

Cohort Retention for Directors

10/30

โœจ What it does

Claude builds a cohort retention screen from [COHORT MONTHS] with logo and revenue curves a director can read in one glance as a working screen you can click in the chat. Once you click a weak cohort, replace the sample numbers with your cohort export.

You are a SaaS finance lead who builds cohort pages a director can read in sixty seconds. <context> I need a cohort retention dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. Board briefing, not a CS ops tool. </context> <inputs> - Company name: [COMPANY NAME] - Cohort months to show: [COHORT MONTHS] - Starting logos and starting ARR per cohort: [OR INVENT] - Retention to plot: [LOGO AND REVENUE] - Horizon: [E.G. 12 MONTHS] </inputs> <task> Build the dashboard with: a KPI row (latest-cohort NRR or month-12 revenue retention, worst cohort, blended logo retention at month 6); a line chart of revenue retention by cohort from [COHORT MONTHS]; a second line or toggle for logo retention; a heatmap table (cohort by month, percent remaining); and a callout of the weakest cohort at month 6 and month 12. Clicking a cohort highlights its line and its heatmap row. Compute remaining percent from labeled arrays; do not hardcode the cards. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Heatmap cells are not color-only; print the percent. - Missing later months show as blank, not as zero retention. - Print-friendly; accessible contrast. </constraints> <format> Return the full HTML file, then explain the retention math and which cohort arrays to replace. </format>

๐Ÿ’ก

Pro tip: Show both logo and revenue. A pretty revenue curve can hide a cohort that already lost half its logos.

Cash

5 prompts

Board Cash and Runway

11/30

โœจ What it does

Claude builds a board cash and runway view from [CASH AND RUNWAY] with a committed versus flexible split and a cut toggle as a working screen you can click in the chat. Click the cash-out month, then replace the sample numbers with bank and burn.

You are a CFO who builds the cash-and-runway board view. <context> I need a board cash and runway view as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. For a board meeting, not a daily treasury desk. </context> <inputs> - Company name: [COMPANY NAME] - Cash, monthly net burn, and runway: [CASH AND RUNWAY] - Committed versus flexible spend this quarter: [AMOUNTS] - Next milestone or raise window: [DATE OR "PATH TO PROFIT"] - Currency: [E.G. USD] </inputs> <task> Build the dashboard with: a KPI row pulled from [CASH AND RUNWAY] (cash, net burn, runway months, cash-out month); a 12-month ending-cash line with a marker on the milestone or raise window; a stacked bar of committed versus flexible spend; a use-of-cash table for the next two quarters; and a short board line that states whether you are raising, cutting, or holding. Include a toggle that cuts 20% of flexible spend and recomputes runway. Clicking the cash-out month marker shows the month name and ending cash. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Runway and cash-out month recompute when the cut toggle is on. - Do not invent a raise that is not in the inputs; if the path is profit, label it that way. - Currency consistent; accessible contrast; print-friendly. </constraints> <format> Return the full HTML file, then explain the cut-toggle math and which cash constants to replace. </format>

๐Ÿ’ก

Pro tip: Split committed payroll and rent from flexible vendor spend before you show a cut. Boards ask what you can stop next month, not what you wish you could.

Thirteen Week Cash Snapshot

12/30

โœจ What it does

Claude builds a thirteen week cash snapshot from [STARTING CASH] with a trough flag at board altitude, not a treasury desk, as a working screen you can click in the chat. After you click the trough week, replace the sample numbers with the cash plan.

You are a CFO who brings a thirteen week cash view into the board pack without turning the meeting into a treasury workshop. <context> I need a thirteen week cash snapshot as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. Board altitude: trough, large items, and ending cash. Not a payment-run desk. </context> <inputs> - Company name: [COMPANY NAME] - Opening cash: [STARTING CASH] - Inflow types: [E.G. COLLECTIONS, NEW SALES, FINANCING] - Outflow types: [E.G. PAYROLL, VENDORS, TAX, RENT] - Large known items with week: [E.G. PAYROLL EVERY OTHER FRIDAY] - Currency: [E.G. USD] </inputs> <task> Build the dashboard with: a KPI row (opening cash from [STARTING CASH], lowest weekly ending balance, week of the trough, net 13-week change); a combined bar-and-line of weekly inflows, outflows, and ending cash; a scenario toggle (base versus late collections that delays 30% of receipts by two weeks); a compact table of the thirteen weeks; and a list of large known items. Ending cash must carry week to week. Flag any week that goes negative. Keep the table scannable from a TV, not a clerk's worksheet. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Scenario toggle recomputes the line, table, trough, and flags from the same constants. - Running balance links week to week; never hardcode an ending cash card. - Currency consistent; accessible contrast; print-friendly. </constraints> <format> Return the full HTML file, then explain the late-collections rule and which week array to replace first. </format>

๐Ÿ’ก

Pro tip: Put payroll and rent on the real week they leave the bank. A flat monthly average hides the week you actually run short.

Burn Path and Profit Date

13/30

โœจ What it does

Claude builds a burn path and profit-date view from [MONTHLY BURN] with a hiring toggle that moves cash-out as a working screen you can click in the chat. Click the hiring toggle, then replace the sample numbers with your real burn.

You are a startup CFO who builds burn-path pages so the board sees cash-out and a path to profit on the same screen. <context> I need a burn path and profit-date dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. Strategy conversation, not a hiring HR tracker. </context> <inputs> - Company name: [COMPANY NAME] - Current cash: [AMOUNT] - Monthly revenue and growth: [AMOUNT AND %] - Current monthly net burn: [MONTHLY BURN] - Planned hires with month and fully loaded cost: [E.G. 1 AE IN MONTH 3 AT $10K] - Horizon: [E.G. 18 MONTHS] </inputs> <task> Build the dashboard with: a KPI row (cash, net burn from [MONTHLY BURN], runway in months, cash-out month, first month net burn is at or below zero if that happens in the horizon); a line of ending cash with and without the planned hires; a hiring toggle that adds or removes those costs and recomputes runway; a table of monthly revenue, costs, net burn, and ending cash; and a callout of how many months the hires consume. Compute runway as cash divided by net burn when burn is positive, and say "not burning" when it is not. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Hiring toggle updates the line, table, runway card, and cash-out month live. - Ending cash carries month to month; cash-out month is the first month ending cash is at or below zero. - Currency consistent; accessible contrast; print-friendly. </constraints> <format> Return the full HTML file, then explain the runway formula and how to edit the hire list. </format>

๐Ÿ’ก

Pro tip: Use fully loaded hire cost, not salary. Benefits, tax, and tools are why runway shrinks faster than the offer letter implies.

Use of Last Raise

14/30

โœจ What it does

Claude builds a use-of-cash view from [RAISE AMOUNT] with planned versus spent buckets since close as a working screen you can click in the chat. Once you click a spend bucket, replace the sample numbers with the raise and spend.

You are a startup CFO who builds a use-of-cash board view after a raise. <context> I need a use-of-cash dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. Board accountability, not an AP clerk screen. </context> <inputs> - Company name: [COMPANY NAME] - Raise amount and close date: [RAISE AMOUNT] - Planned buckets and amounts: [E.G. PRODUCT 40%, GTM 35%, RUNWAY BUFFER 25%] - Spent to date by bucket: [AMOUNTS OR INVENT] - Months since close: [NUMBER] - Currency: [E.G. USD] </inputs> <task> Build the dashboard with: a KPI row (raised from [RAISE AMOUNT], spent, remaining, months of announced plan left); a stacked bar of planned versus spent by bucket; a remaining-cash line since the close date; a table of buckets with plan, spent, remaining, and percent used; and a callout of the bucket that is running hottest versus the announced mix. Clicking a bucket filters the table and the callout. Remaining cash must equal raise minus spent across buckets, plus any leftover not allocated. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Totals reconcile to the raise; flag a break if spent exceeds a bucket plan. - Bucket click links bar, table, and callout. - Currency consistent; accessible contrast; print-friendly. - Do not invent a new raise or a valuation. </constraints> <format> Return the full HTML file, then explain how to replace the raise, the bucket plan, and spent-to-date. </format>

๐Ÿ’ก

Pro tip: Use the bucket names from the investor update, not a new taxonomy. The page is there to show you are spending the raise you already described.

Minimum Cash and Covenant

15/30

โœจ What it does

Claude builds a minimum-cash and covenant screen from [MINIMUM CASH] with headroom and a breach month as a working screen you can click in the chat. First click the covenant line, then replace the sample numbers with your policy and balances.

You are a CFO who builds minimum-cash and covenant pages so the board sees headroom before a breach. <context> I need a minimum cash and covenant dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. Policy and headroom, not a multi-account clerk view. </context> <inputs> - Company name: [COMPANY NAME] - Minimum cash policy or covenant: [MINIMUM CASH] - Current cash and a 12-month ending-cash path: [OR INVENT] - Other covenants if any: [E.G. DEBT SERVICE, MINIMUM RUNWAY MONTHS, NONE] - Currency: [E.G. USD] </inputs> <task> Build the dashboard with: a KPI row (cash, minimum from [MINIMUM CASH], headroom $, months of headroom, nearest breach month if any); a 12-month ending-cash line with a policy line; a headroom area between cash and the floor; a covenant table with limit, actual, headroom, and pass/fail; and a callout of the first month the sample path would fail. Clicking a covenant row focuses that limit on the chart when it is a cash floor, or highlights the table row otherwise. Compute headroom as cash minus the floor. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Pass/fail is computed from the constants; do not invent a lender name I did not give. - Months under the floor flagged with color and the word breach. - Currency consistent; accessible contrast; print-friendly. - Label this as a tracker, not legal advice on a credit agreement. </constraints> <format> Return the full HTML file, then explain the headroom math and which cash and limit constants to replace. </format>

๐Ÿ’ก

Pro tip: Put the real covenant or board policy number on the page. A pretty cash line with no floor cannot show headroom.

These prompts give you the what. Tutorials give you the why.

Learn when to use extended thinking, how to build Claude Projects, and workflows that compound. 300+ tutorials and growing.

Try AI Academy Free

OKRs

5 prompts

Company OKR Scorecard

16/30

โœจ What it does

Claude builds a company OKR scorecard from [OKR LIST] with objective RAG, KR bars, and a decision callout as a working screen you can click in the chat. After you click a red objective, replace the sample numbers with this quarter's OKRs.

You are a chief of staff who builds company OKR scorecards for a board, not a team standup. <context> I need a company OKR scorecard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. Company objectives only. No sprint boards. </context> <inputs> - Company name: [COMPANY NAME] - Quarter: [E.G. Q3 2026] - Objectives and key results with start, current, and target: [OKR LIST] - RAG rule: [E.G. GREEN AT OR ABOVE 0.7, AMBER AT OR ABOVE 0.4, ELSE RED] - Owner per objective: [NAMES] </inputs> <task> Build the dashboard with: a header for the quarter; a KPI row (objectives green / amber / red, average KR score, count of KRs at zero); a card per objective from [OKR LIST] with owner, RAG, and a bar per KR (current versus target, score 0-1); a table of all KRs sorted worst first; and a callout of the one objective that needs a board decision. Compute KR score as current / target when the KR is a maximize, and as a stated 0-1 if I give a score. Clicking an objective expands its KRs and dims the others. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Objective RAG is computed from its KR scores against [RAG RULE], not hardcoded. - Status uses color plus the word. Print-friendly. Accessible contrast. - Do not invent a fourth objective. Do not add Jira or sprint widgets. </constraints> <format> Return the full HTML file, then explain the scoring rule and which OKR array to replace. </format>

๐Ÿ’ก

Pro tip: Score key results with a number, not a paragraph. If a KR cannot move a bar, it is a slogan and the board will treat it as one.

Objective RAG Grid

17/30

โœจ What it does

Claude builds an objective RAG grid from [OBJECTIVES] with owners, scores, and a count banner as a working screen you can click in the chat. Click an amber objective, then replace the sample numbers with the real scores.

You are a chief of staff who builds a one-page objective RAG grid for directors. <context> I need an objective RAG grid as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. Board page, not a team wiki. </context> <inputs> - Company name: [COMPANY NAME] - Objectives with owner, score 0-1, and one-line status: [OBJECTIVES] - RAG thresholds: [E.G. 0.7 / 0.4] - Quarter: [E.G. Q3 2026] </inputs> <task> Build the dashboard with: a count banner (green, amber, red); a grid of objective cards from [OBJECTIVES] showing name, owner, score, RAG word, and the one-line status; a simple bar of scores; and a click that opens a footnote with the scoring rule. Filter chips for red only and amber plus red. Compute RAG from the score and the thresholds. Keep it to one screen at laptop width. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Banner counts recompute from the objectives array. - Color plus word; print-friendly; accessible contrast. - No key-result editor, no check-ins, no chat. Grid only. </constraints> <format> Return the full HTML file, then restate the thresholds and which array to replace. </format>

๐Ÿ’ก

Pro tip: One row per company objective. If you paste every team OKR, the grid becomes an ops floor and the board will stop reading.

Key Result Pace Chart

18/30

โœจ What it does

Claude builds a key result pace chart from [KEY RESULTS] with elapsed fraction and implied finish as a working screen you can click in the chat. Once you click an off-pace KR, replace the sample numbers with current KR values.

You are an operator-turned-chief-of-staff who builds KR pace charts a board can read. <context> I need a key result pace dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. Company KRs, not a sprint burndown. </context> <inputs> - Company name: [COMPANY NAME] - Key results with start, current, target, and type (flow or binary): [KEY RESULTS] - Elapsed fraction of the quarter: [E.G. 0.55] - Quarter: [E.G. Q3 2026] </inputs> <task> Build the dashboard with: a KPI row (KRs on pace, off pace, binary done, binary not done); a bar of current versus a straight-line expected value (start + (target-start) * elapsed) for flow KRs; a table from [KEY RESULTS] with current, expected, implied finish, and on/off-pace; and a callout of the furthest-behind flow KR. Binary KRs show done or not done and are never annualized. Clicking a KR highlights its bar and shows the pace math in a footnote. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Pace is computed from start, current, target, and elapsed fraction. - Off-pace flagged with color and the word off-pace. Guard divide-by-zero. - Print-friendly; accessible contrast. </constraints> <format> Return the full HTML file, then explain the expected-value formula and which KR array to replace. </format>

๐Ÿ’ก

Pro tip: Mark binary KRs as stock, not a run rate. Annualizing "shipped" or "not shipped" is how a pack lies at mid-quarter.

Commit Versus Stretch

19/30

โœจ What it does

Claude builds a commit versus stretch OKR pack from [COMMIT LINE] so the board sees what was promised versus what was hoped as a working screen you can click in the chat. Click a missed commit, then replace the sample numbers with commit and stretch scores.

You are a CEO who separates board commits from stretch so a miss cannot hide in a hope. <context> I need a commit versus stretch OKR pack as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. Governance, not a team pep wall. </context> <inputs> - Company name: [COMPANY NAME] - Commit objectives and KRs with current and target: [COMMIT LINE] - Stretch objectives and KRs: [OR NONE] - Quarter: [E.G. Q3 2026] - RAG rule for commits: [E.G. GREEN AT 1.0, AMBER AT 0.7, ELSE RED] </inputs> <task> Build the dashboard with: two columns (Commit and Stretch); a KPI row that scores commits only for the banner the board should care about; cards from [COMMIT LINE] with RAG; stretch cards visually quieter; a table of missed commits; and a callout of the one commit that needs a decision. A toggle hides stretch entirely. Clicking a missed commit expands the KR bars. Do not let stretch scores lift the commit banner. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Commit banner ignores stretch scores. - Missed commits flagged with color and the word missed. - Print-friendly; accessible contrast. </constraints> <format> Return the full HTML file, then explain why stretch is excluded from the banner and which two arrays to replace. </format>

๐Ÿ’ก

Pro tip: Label commits in the same words you used when the board approved the quarter. Renaming a miss as a stretch after the fact will get caught.

Cross-Team OKR Map

20/30

โœจ What it does

Claude builds a cross-team OKR map from [TEAM OKRS] with blocked KRs and a dependency list as a working screen you can click in the chat. First click a blocked KR, then replace the sample numbers with each team's OKRs.

You are a chief of staff who maps company OKR dependencies so the board sees what is blocked, not who is busy. <context> I need a cross-team OKR map as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. Company-level dependencies only. </context> <inputs> - Company name: [COMPANY NAME] - Teams and their top objectives or KRs: [TEAM OKRS] - Known dependencies: [E.G. GTM KR BLOCKED ON PRODUCT KR] - Quarter: [E.G. Q3 2026] </inputs> <task> Build the dashboard with: a KPI row (teams, KRs, blocked count, unique blockers); a simple map or grouped list of [TEAM OKRS] with lines or rows for each dependency; a blocked table (KR, owner team, blocker team, blocker KR, status); and a callout of the blocker that unlocks the most KRs. Clicking a blocked KR highlights its blocker and the other KRs waiting on it. Compute blocked count from the dependency list, not a hardcoded card. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Do not invent a dependency I did not list. - Blocked rows flagged with color and the word blocked. - Print-friendly; accessible contrast; readable on a laptop, not a 4K ops wall. </constraints> <format> Return the full HTML file, then explain how to replace the team KR list and the dependency list. </format>

๐Ÿ’ก

Pro tip: Only map dependencies the board already heard about. A web of every team KR is an ops floor dressed as strategy.

Risks

5 prompts

Top Board Risk Register

21/30

โœจ What it does

Claude builds a top board risk register from [RISK LIST] with likelihood, impact, owner, and next review as a working screen you can click in the chat. After you click the hottest risk, replace the sample numbers with the live register.

You are a chief of staff who builds a top-ten risk register for a board meeting. <context> I need a top board risk register as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. Directors, not an incident floor. </context> <inputs> - Company name: [COMPANY NAME] - Risks with likelihood 1-5, impact 1-5, owner, trend, next review: [RISK LIST] - Score rule: [E.G. LIKELIHOOD TIMES IMPACT] - Board date: [BOARD DATE] </inputs> <task> Build the dashboard with: a KPI row (risk count, red count, highest score, overdue reviews); a ranked table from [RISK LIST] with name, category, L, I, score, owner, trend, next review; a bar of scores; and a click that opens a one-line description plus the last mitigation note (sample text). Sort hottest first. Compute score from the rule. Flag overdue reviews against [BOARD DATE]. Trend shown as rising, stable, or falling with a word, not an icon only. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Scores and overdue flags are computed, not hardcoded. - Do not invent legal conclusions. This is a register, not counsel. - Print-friendly; accessible contrast; max about ten rows on the first screen. </constraints> <format> Return the full HTML file, then explain the score rule and which risk array to replace. </format>

๐Ÿ’ก

Pro tip: Cap the register at the ten risks the board already owns. A fifty-row ops log will not get read, and the real red will hide.

Likelihood Impact Heat Map

22/30

โœจ What it does

Claude builds a likelihood-impact heat map from [RISK CATEGORIES] with clickable cells and a ranked list as a working screen you can click in the chat. Click a high-high cell, then replace the sample numbers with scored risks.

You are a risk lead who builds a likelihood-impact heat map a board can point at. <context> I need a likelihood-impact heat map as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. Board page, not a SOC wall. </context> <inputs> - Company name: [COMPANY NAME] - Risks with likelihood, impact, and category: [RISK CATEGORIES] - Axis scale: [E.G. 1-5] - Red zone: [E.G. SCORE AT OR ABOVE 15, OR L AND I BOTH AT OR ABOVE 4] </inputs> <task> Build the dashboard with: a 5 by 5 (or stated scale) heat map that plots each risk from [RISK CATEGORIES]; cell counts; a ranked list beside the map; category chips that dim other risks; and a click on a cell that lists the risks in that cell. Compute score and red-zone membership from the constants. Print risk initials or short names in the cell when a cell has one or two risks, and a count when it has more. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Heat is not color-only; print the count or score class (low, mid, high). - Category chips must filter. Print-friendly; accessible contrast. - Do not invent a residual score I did not give. </constraints> <format> Return the full HTML file, then explain the red-zone rule and which risk array to replace. </format>

๐Ÿ’ก

Pro tip: Keep the axes at 1-5 the way your last pack did. A new 1-3 scale mid-year makes last quarter's red look like a different sport.

Mitigation Status Tracker

23/30

โœจ What it does

Claude builds a mitigation status tracker from [MITIGATIONS] with due dates, percent done, and overdue flags as a working screen you can click in the chat. Once you click an overdue action, replace the sample numbers with the mitigation list.

You are a chief of staff who tracks mitigations the board already asked for. <context> I need a mitigation status tracker as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. Actions the board can see, not a Jira board. </context> <inputs> - Company name: [COMPANY NAME] - Mitigations with risk name, owner, due date, percent done, status: [MITIGATIONS] - Board date used as today: [BOARD DATE] - Statuses: [E.G. NOT STARTED, ON TRACK, AT RISK, DONE, OVERDUE] </inputs> <task> Build the dashboard with: a KPI row (open, overdue, percent of actions done, risks with no action); a bar of status counts; a table from [MITIGATIONS] with risk, action, owner, due, percent, status; and a click that shows a one-line next step (sample). Overdue is computed when due date is before [BOARD DATE] and status is not done. A chip filter for overdue and at risk. Sort overdue first, then nearest due date. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Overdue is computed, not a hardcoded label. - Status uses color plus the word. Print-friendly; accessible contrast. - Do not add time-tracking or story points. </constraints> <format> Return the full HTML file, then explain the overdue rule and which mitigation array to replace. </format>

๐Ÿ’ก

Pro tip: One action per risk on this page. A task dump belongs in the ops tool, not in the board packet.

Red Risk Deep Dive

24/30

โœจ What it does

Claude builds a red-risk deep dive from [RED RISKS] with residual score, trigger, and a board ask as a working screen you can click in the chat. Click the residual score, then replace the sample numbers with the risks you will defend.

You are a CEO who deep-dives only the red risks so the board can decide, not browse. <context> I need a red-risk deep dive as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. Two to five reds, not the full register. </context> <inputs> - Company name: [COMPANY NAME] - Red risks with inherent score, residual score, trigger, mitigation, and ask: [RED RISKS] - Ask types: [E.G. WATCH, SPEND, DECIDE, ACCEPT] - Board date: [BOARD DATE] </inputs> <task> Build the dashboard with: a header that says these are the reds only; a card per risk from [RED RISKS] with inherent versus residual, trigger, mitigation, owner, and the ask; a small bar of residual scores; and a click that expands the trigger and the ask into a short board paragraph (sample you can edit). A filter by ask type. Residual must be less than or equal to inherent; flag a break if a sample row violates that. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Do not invent legal, insurance, or valuation advice. - Ask type printed as a word. Print-friendly; accessible contrast. - No incident timeline unless I put dates in the inputs. </constraints> <format> Return the full HTML file, then list which residual and ask fields to replace. </format>

๐Ÿ’ก

Pro tip: Every red risk needs an ask (watch, spend, decide, accept). A red with no ask is a slide, not governance.

Legal and Compliance Watch

25/30

โœจ What it does

Claude builds a legal and compliance watch from [COMPLIANCE ITEMS] with status, next date, and residual exposure as a working screen you can click in the chat. First click an open item, then replace the sample numbers with counsel's list.

You are a chief of staff who builds a legal and compliance watch for the board. You track items. You do not give legal advice. <context> I need a legal and compliance watch as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. A banner must say this is a tracker, not counsel. </context> <inputs> - Company name: [COMPANY NAME] - Items with type, status, next date, owner, residual exposure if known: [COMPLIANCE ITEMS] - Types: [E.G. PRIVACY, EMPLOYMENT, COMMERCIAL, IP, OTHER] - Board date: [BOARD DATE] </inputs> <task> Build the dashboard with: a banner that this is a tracker I will confirm with counsel; a KPI row (open items, next date, items past [BOARD DATE], items with unknown exposure); a table from [COMPLIANCE ITEMS]; a bar by type; and a click that opens the next date and a one-line note. If residual exposure is unknown, print unknown, do not invent a dollar. Sort by next date. Flag past-due dates as late on the calendar only. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Do not invent a statute, a fine, a regulator, or a filing outcome. - Unknown exposure stays unknown. Print-friendly; accessible contrast. - No legal conclusion in the callout. </constraints> <format> Return the full HTML file, then list the fields I must replace and repeat that counsel has to review the items. </format>

๐Ÿ’ก

Pro tip: Paste counsel's item names. Do not let Claude invent a statute, a fine, or a filing date you did not give.

Most people use 10% of Claude. Tutorials unlock the rest.

AI Academy: 300+ hands-on tutorials on Claude, ChatGPT, Midjourney, and 50+ AI tools. New tutorials added every week.

Start Your Free Trial

One Pager

5 prompts

Single Screen Board One-Pager

26/30

โœจ What it does

Claude builds a single-screen board one-pager dated [PACKET DATE] with KPI, ARR, cash, OKRs, and risks as a working screen you can click in the chat. After you click a section, replace the sample numbers with the full pack figures.

You are a CFO who builds a single-screen board one-pager that can stand in for the pack when time is short. <context> I need a single-screen board one-pager as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. For a board meeting, not an ops floor. No scroll on a 13-inch laptop if you can help it. </context> <inputs> - Company name: [COMPANY NAME] - Packet date: [PACKET DATE] - Eight KPI actuals and plans: [OR INVENT] - ARR beginning, movements, ending: [OR INVENT] - Cash, burn, runway: [OR INVENT] - Three company OKRs with scores: [OR INVENT] - Three top risks with scores: [OR INVENT] - One-line narrative: [OR DRAFT FROM THE NUMBERS] </inputs> <task> Build one page dated [PACKET DATE] with five tight sections: KPI strip (eight cards, RAG computed); ARR mini-bridge; cash and runway; three OKR scores; three risks. Header holds the one-line narrative. Clicking a section expands a footnote with the definition. Compute RAG, runway, and ARR ending from the constants. Keep type large enough to read from a TV at the end of a table. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - No extra pages, no appendix, no ticket widgets. - Color plus word for RAG. Print-friendly light background. Accessible contrast. - Do not invent a raise or a valuation. </constraints> <format> Return the full HTML file, then list which constants to replace so every section still ties. </format>

๐Ÿ’ก

Pro tip: Five sections, one screen. If you need a scroll to see cash, it is no longer a one-pager and the board will not print it.

CEO Pre-Read Sheet

27/30

โœจ What it does

Claude builds a CEO pre-read sheet around [CEO NARRATIVE] with eight numbers and three footnotes as a working screen you can click in the chat. Click a footnote, then replace the sample numbers so the letter still ties.

You are a CEO writing a pre-read the board can finish in four minutes. <context> I need a CEO pre-read sheet as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. A letter plus numbers. Not a blog and not an ops report. </context> <inputs> - Company name: [COMPANY NAME] - Board date: [BOARD DATE] - The story I will stand behind: [CEO NARRATIVE] - Eight numbers with labels and plans: [OR INVENT] - Three footnotes I must not skip: [E.G. DEFINITION, ONE-TIME ITEM, RISK] - What I will not claim: [OFF-LIMITS] </inputs> <task> Build the dashboard with: a left column of three short paragraphs built from [CEO NARRATIVE] that cite the eight numbers by name; a right column of the eight cards with actual, plan, and RAG; three numbered footnotes; and a click on a footnote that highlights the card it belongs to. If a sample number would break the narrative, flag that sentence. Honor [OFF-LIMITS]. Keep the whole sheet to one printed page at A4 or letter. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Do not invent a customer, a raise, or a metric I did not give. - Print-friendly; accessible contrast; page-break safe. </constraints> <format> Return the full HTML file, then list which sentences cite which cards and which constants to replace. </format>

๐Ÿ’ก

Pro tip: The letter is three short paragraphs max. If the numbers cannot support a sentence, cut the sentence before the meeting, not after a director asks.

Print Pack Cover

28/30

โœจ What it does

Claude builds a print-friendly pack cover from [MEETING AGENDA] with time boxes, owners, and the five numbers on the table as a working screen you can click in the chat. Once you click an agenda item, replace the sample numbers with this meeting's agenda.

You are a board secretary who builds a print pack cover the room can follow. <context> I need a print-friendly pack cover as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. Cover page for a board meeting, not a conference schedule. </context> <inputs> - Company name: [COMPANY NAME] - Meeting date, time, and place or video: [MEETING META] - Agenda items with minutes and owner: [MEETING AGENDA] - Five numbers the room should see before item one: [OR INVENT] - The ask if any: [OR NONE] </inputs> <task> Build the cover with: a header (company, date, closed session flag if I say so); five KPI chips; an agenda table from [MEETING AGENDA] with item, owner, minutes, and a running clock that sums to the meeting length; a box for [THE ASK] or a quiet "no formal ask" if NONE; and a click on an agenda row that highlights that item and shows its minute share of the meeting. Compute total minutes and flag if the sample agenda exceeds the stated meeting length. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Running clock is computed from the minute column. - Print-friendly; light background; works in grayscale. - No marketing fluff. No ops floor widgets. </constraints> <format> Return the full HTML file, then explain the minute math and which agenda array to replace. </format>

๐Ÿ’ก

Pro tip: Put a real minute count on each item. An agenda without clocks is how cash gets eight minutes and a product demo gets forty.

Ask of the Board

29/30

โœจ What it does

Claude builds an ask-of-the-board page from [THE ASK] with options, cash impact, and a recommended vote as a working screen you can click in the chat. Click an option card, then replace the sample numbers with the real ask.

You are a CEO putting a single ask in front of the board. <context> I need an ask-of-the-board page as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. One decision. Not a strategy essay. </context> <inputs> - Company name: [COMPANY NAME] - The ask in one sentence: [THE ASK] - Options (including do nothing): [E.G. APPROVE, APPROVE A SMALLER NUMBER, DEFER, REJECT] - Cash or runway impact per option: [OR INVENT PLAUSIBLE ONES] - Recommendation: [WHICH OPTION] - What happens if we wait one quarter: [COST OF WAIT] </inputs> <task> Build the dashboard with: a header that prints [THE ASK]; option cards with cash impact, runway change, and a one-line tradeoff; a recommended badge on the option I named; a small table of impacts; and a "if we wait" callout from [COST OF WAIT]. Clicking an option highlights it as the working vote and updates a summary line. Compute runway change from the cash constants. Do not invent a valuation or a legal finding. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Impacts recompute from labeled constants. - Print-friendly; accessible contrast; one screen. - Do not add a second ask. </constraints> <format> Return the full HTML file, then list which impact constants to replace before you send the pack. </format>

๐Ÿ’ก

Pro tip: State the vote in one sentence (approve X up to $Y through DATE). If the ask is a vibe, you will get a vibe back.

Decision One-Slide

30/30

โœจ What it does

Claude builds a decision one-slide from [THE DECISION] with context, options, and a residual risk line as a working screen you can click in the chat. First click the recommended option, then replace the sample numbers with the decision memo.

You are a CFO who compresses a decision memo onto one slide the board can vote from. <context> I need a decision one-slide as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic sample data so it opens as a working screen I can click in the chat. One vote. Not a 20-page memo and not an ops runbook. </context> <inputs> - Company name: [COMPANY NAME] - Decision in one line: [THE DECISION] - Context in three bullets I can prove: [CONTEXT] - Options with cost, upside, residual risk: [OR INVENT PLAUSIBLE ONES] - Recommended option: [NAME] - Deadline: [DATE] </inputs> <task> Build one slide with: a title from [THE DECISION]; three context bullets; option cards (cost, upside, residual risk score if given); a recommended outline; a residual-risk line that updates when I click an option; and a deadline chip. Clicking an option makes it the working choice and rewrites a one-line vote text ("Vote: [option] by [DATE]"). Keep it to one screen at 1280px wide and one printed landscape page. </task> <constraints> - One self-contained HTML file; only a charting library from a CDN is allowed. - Do not invent a number I did not give; mark missing upside as unknown. - Print-friendly; accessible contrast; large type. - No appendix link farm. The slide is the memo. </constraints> <format> Return the full HTML file, then list which option fields to replace so the vote line still matches. </format>

๐Ÿ’ก

Pro tip: If the residual risk does not change by option, you do not have options. You have a narrative with buttons.

Free tool

Prompt Optimizer

Turn a rough idea into a structured, professional AI prompt.

Try it free โ†’

Frequently Asked Questions

Yes. Each prompt asks Claude to return one self-contained HTML file with inline CSS, vanilla JS, and a charting library from a CDN, already filled with sample numbers. It opens as a working screen you can click in the chat: cards, toggles, tables, and charts respond before you paste the real pack.
Finance dashboards go deep on P and L, close, and household money. Ops dashboards run a floor: queues, OTIF, incidents. Sales dashboards run a book: pipeline, activity, forecast. This page is the board meeting: KPI pack, ARR, cash, OKRs, risks, and a one-pager a director can read from a TV. Use the others to operate. Use this one to brief.
Every prompt keeps sample data in labeled JS constants near the top of the file. Click through the sample first so you know which control a director will poke, then replace those arrays with the pack you will defend. Cards, RAG, runway, bridges, and OKR scores recompute from the constants instead of from hardcoded labels.
Board meeting. The prompts refuse warehouse widgets, ticket queues, and rep activity boards on purpose. If you need a floor screen, use the ops or sales dashboard pages. If you need the books, use finance dashboards. If you need one page the room can vote from, stay here.
Yes. Copy the HTML out of the chat, save it as a file, and open it in a browser, drop it on any static host, or print the one-pagers. It is one file, so a board packet does not need a server. For live data later, use Claude's closing note on swapping the sample arrays for a fetch or a CSV.

Prompts are the starting line. Tutorials are the finish.

A growing library of 300+ hands-on tutorials on ChatGPT, Claude, Midjourney, and 50+ AI tools. New tutorials added every week.

7-day free trial. Cancel anytime.