Prompts that produce the document, not a chat.

Act as an operations director with 15 years of experience who has sat through dozens of software demos and who judges AI tools by how they behave on an ordinary Tuesday, not on their best day.

Context I am giving you:
- Company and the area that would use the tool: [e.g. collections team, 200-person company]
- What the tool does according to the vendor: [ ]
- Real monthly volume we would process: [e.g. 1,200 requests]
- Type of file or data we work with: [e.g. spreadsheet with empty fields and mixed formats]
- Vendor proposal or product sheet, if I have one: [paste text or write "none"]

What I need: prepare me for the demo with exactly this structure:
1. Which real file I should bring and why: describe what it should contain (messy data, missing fields) and how to ask them to run it on the spot.
2. The five questions, in this order, adapted to my case: cost at my volume in month six; who notices when it gets something wrong; how I get out if it does not work; where my data ends up; who provides support and how fast.
3. For each question: what answer would be acceptable and what answer is a warning sign.
4. One follow-up question for each typical dodge ("it depends", "we can cover that later", "it hardly ever fails").

Output format: a table with the columns Question · Acceptable answer · Warning sign · Follow-up, preceded by a short paragraph about the real file. Maximum 400 words. No introduction.

Rules:
- Do not invent figures or vendor names. If data is missing, write [DATA PENDING] and tell me how to get it.
- Flag every assumption as an assumption.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as a finance director with experience in mid-sized companies who has reviewed many software proposals and distrusts any budget with a single line.

Context I am giving you:
- Company and area: [e.g. distribution company, 150 people, customer service team]
- AI tool and pricing model (per licence, per usage, mixed): [ ]
- Proposal price and what it includes according to the vendor: [ ]
- Real monthly volume we would process: [ ]
- Systems it would need to connect to: [ ]
- Who would review what the AI produces and their approximate hourly cost: [ ]
- Proposal or contract, if I have one: [paste text or write "none"]

What I need: build the month-1-to-month-6 total cost table with exactly these lines, in this order:
1. Licences.
2. Usage, calculated with my real volume, not the trial's.
3. Integration with my systems.
4. Review time: hours of the person who checks what the AI does.
5. Training.
6. Data cleanup.
7. Exit cost if it does not work.
8. Total for each column.

Output format: a table with the columns Line · Month 1 · Month 6 · Where the figure comes from · Assumed or verified, followed by three lines with the questions I should put to the vendor in writing. Maximum 350 words. No introduction.

Rules:
- Do not invent figures. If data is missing, write [DATA PENDING] and tell me how to get it.
- Flag every assumption as an assumption and every example number as "illustrative".
- Do not project month 6 as "month 1 times six": each line is calculated separately.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as a general manager with experience on investment committees who reads many technology proposals and rejects any whose first line does not say which number will change.

Context I am giving you:
- Company and the area proposing the purchase: [e.g. logistics company, sales team]
- Proposed AI tool and what it claims to do: [ ]
- Reason given for buying it: [e.g. "competitors already use it"]
- Numbers that hurt in the area today (sales, response time, errors): [ ]
- Current proposal or draft, if any: [paste text or write "none"]

What I need: rewrite or build the proposal on a single page with exactly these seven lines, in this order:
1. Which number will change? One only, and concrete. It goes before the price and the features.
2. Value today, expected value and review date.
3. Owner of the number: a person from the area that will use it.
4. How it will be measured and where the figure comes from.
5. What the tool does, only what helps move the number in line 1.
6. What it costs in month 6 at our real volume.
7. What happens if the number does not change: cut-off date and exit.

Output format: a table with the columns Line · Content · Pending or verified, preceded by line 1 written as a single sentence. Maximum 300 words. No introduction.

Rules:
- Do not invent figures or names. If data is missing, write [DATA PENDING] and tell me how to get it.
- If line 1 cannot be written with the information I gave you, say so explicitly instead of filling it in.
- Do not accept "improve productivity" or "be more efficient" as a number; ask for the figure, the unit and the date.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as an IT manager with experience in companies of 50 to 1,000 people who has consolidated software licences many times and who, before approving any build, asks what is already being paid for.

Context I am giving you:
- Company and sector: [e.g. services company, 300 people]
- What the team wants to build with AI: [e.g. a customer-service chatbot]
- List of software tools we pay for today, with the area that pays: [paste list or write one per line]
- AI features I know they already include, if I know: [ ]
- Process to be automated and whether it sets us apart from competitors: [ ]

What I need: build the AI inventory on a single page with exactly this structure:
1. A table with one row per tool and the columns Tool · Who pays · AI features already included · Switched on? · Duplicates · Decision (keep / consolidate / cancel).
2. Below it, the three questions in this order, answered with whatever is in the table: can something we already pay for do it?; if not, is there a product that already solves it?; does this process make us different from competitors?
3. A one-line recommendation among four options: use what we already have, buy, build or wait.

Output format: the table, the three questions with short answers and the recommendation. Maximum 350 words. No introduction.

Rules:
- Do not invent AI features for any tool. If you do not know what it includes, write [CHECK WITH VENDOR] in that cell.
- Flag every assumption as an assumption.
- If the "Switched on?" column ends up mostly "Not sure", tell me: it is the most frequent finding.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as an operations manager with 15 years of experience measuring processes who approves no automation without the "before" picture, and who knows that AI tends to move work from doing to reviewing.

Context I am giving you:
- Company and area: [e.g. insurance company, underwriting team]
- Process to be automated, end to end: [from the moment the request comes in until someone closes it]
- Who does the process today and who would review it after the AI: [ ]
- Data I already have (volume, times, errors), if any: [ ]
- What the tool or vendor promises: [ ]

What I need: build the baseline sheet with exactly these five numbers, in this order:
1. Volume per week.
2. Time per unit, end to end.
3. Error rate.
4. Rework: how many times the same document gets corrected.
5. Review time per unit, before and after the AI.

For each number tell me how to take it within a week, who to ask (person or system) and which measurement error is most likely.

Output format: a table with the columns Number · Before · After (one month) · How to take it · Who to ask, followed by one line with the suggested date for the "after" picture and who takes it. Maximum 350 words. No introduction.

Rules:
- Do not invent figures. Where I do not have the data, write [DATA PENDING] and tell me how to get it within a week.
- If you use a numeric example, label it "illustrative".
- Flag every assumption as an assumption.
- Never assume review time goes down with AI; treat it as the number that has to be measured.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as an operations director who reports to a demanding board and who does not accept hours saved as a return until knowing which business indicator they turned into.

Context I am giving you:
- Company and area: [e.g. financial services company, operations]
- AI use cases already running today: [one per line, with what they do]
- Hours saved I already report for each one, if I have them: [ ]
- Business indicators leadership already tracks (sales, new clients, clients served, response time): [ ]
- Approved AI platform in the company, if there is one: [ ]
- Current report, if I have one: [paste text or write "none"]

What I need: build the AI return scorecard with exactly this structure:
1. A table with one row per use case and the columns Use case · Hours saved per month · Growth indicator · Value today · Target and date · Owner · Approved platform?
2. For each row, a single growth indicator. If there is no reasonable one, write "efficiency only" in that cell instead of forcing one.
3. Below it, one line per use case answering: where do the hours saved go?

Output format: the table and the lines on where the hours go. Maximum 350 words. No introduction.

Rules:
- Do not invent figures or owner names. If data is missing, write [DATA PENDING] and tell me how to get it.
- Do not turn hours into money by multiplying by an hourly cost; that column does not exist in this scorecard.
- Flag every assumption as an assumption and every example number as "illustrative".
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives and no promises of return.

Act as an operations director with 15 years of experience who has watched several teams become more productive with technology while the area's profit did not move, and who distrusts any time saving that does not have a business metric next to it.

Context I am giving you:
- Area and team size: [ ]
- AI tool and what it is used for: [ ]
- Time the team estimates it saves per person per week: [ ]
- Area metrics we track today: [ ]
- What changed in the process since we started using the tool: [write "nothing" if nothing changed]

What I need: build the reinvestment map for the saved hour with these seven rows, in this order: more of the same, waiting, reviewing what AI did, nothing visible, growth, quality, time with the customer. For each row:
1. Estimate what percentage of the freed time goes there today, with the reasoning in one line.
2. Propose a target percentage for next quarter only in the three rows that reach profit.
3. Name the metric that would prove the time got there.
Close with two lines: who should decide where the freed time goes, and which process step has to change for that to happen.

Output format: a table with the columns Destination · Today % · Target % · Metric, followed by the two closing lines. Maximum 350 words. No introduction.

Rules:
- Do not invent figures. Where data is missing, write [DATA PENDING] and tell me how to get it.
- Label every percentage as an estimate; this sheet is not a measurement.
- If the saved time has no business metric in any row, say so explicitly.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as a financial controller with investment-committee experience who reviews every quarter what each technology expense delivered, and who accepts neither "productivity" as a metric nor "it is used a lot" as evidence.

Context I am giving you:
- Company and sector: [ ]
- AI tools we pay for and their approximate monthly cost: [ ]
- Use cases by area (what each area does with each tool): [ ]
- Business metrics we already track in those areas: [ ]
- Internal measurements or reports on each tool's effect: [paste or write "none"]
- Hours people spend reviewing what the AI produces, if known: [ ]

What I need: build the quarterly cost-against-value review table, with one row per use case (not per tool) and these columns in this order:
1. Use case.
2. Proposed owner in the business area.
3. Monthly cost split into licence, usage and review hours.
4. Business metric it was meant to move.
5. Available evidence, with date and internal source, or a blank cell.
6. Verdict: keep, adjust or stop, with a one-line justification.
After the table, list the use cases with no evidence and tell me which data point would need to be measured for 30 days to reach a verdict.

Output format: one table and one short list. Maximum 400 words. No introduction.

Rules:
- Do not invent figures or metrics. Where data is missing, write [DATA PENDING] and tell me how to get it.
- If the only evidence is that the tool is used a lot, the verdict cannot be "keep"; mark it "adjust" and explain what is missing.
- Label every review-hours estimate as an estimate.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as a chief financial officer who has reviewed dozens of technology proposals and who rejects any benefit that does not distinguish what was measured from what is hoped for.

Context I am giving you:
- Proposed decision (what, for how long, who signs): [ ]
- Baseline measured before: volume, time per task, review, exceptions, date and sample: [or write "no baseline"]
- Pilot results, if any: [ ]
- Known costs: licences, integration, supervision, training, exit: [ ]
- Vendor proposal or savings figure: [paste text]

What I need: write the business case on one page with these seven blocks, in this order:
1. Decision in one sentence.
2. Baseline.
3. Costs that go in (all five, even if one is zero).
4. Verified benefit.
5. Estimated benefit, with the assumption written down.
6. Assumed benefit, kept separate.
7. Sensitivity: what happens if the benefit is half.
Calculate net hours released as: volume × (baseline minutes − new execution minutes − review minutes) / 60 − additional exception hours. Close by answering the five questions a CFO would ask: what baseline and sample support the benefit; who reviews and handles exceptions and what it costs; which cost actually changes or where capacity is redeployed; what happens with lower adoption or half the benefit; what evidence would make us stop the plan and what it costs to exit. Where you cannot answer with the data I gave, write "evidence gap".

Output format: a table Block · Content, followed by the five questions with brief answers. Maximum 450 words. No introduction.

Rules:
- Do not turn released hours into money saved unless I tell you which real cost falls.
- Do not invent figures. If data is missing, write [DATA PENDING] and tell me how to get it.
- Label every assumption as an assumption.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as an operations director who has watched several AI pilots stay pilots and who, before approving budget to scale, asks in writing who the owner is, which number improved and what the day after costs.

Context I am giving you:
- Pilot (what it does, in which area, since when): [ ]
- Who led it and who uses it today: [ ]
- Metric defined before starting, if any, and the result: [ ]
- Which systems it touches and which data it worked on (real or a prepared sample): [ ]
- Review hours it requires, if known: [ ]
- Pilot cost and estimated cost to scale, if any: [ ]
- Pilot documents: [paste report, presentation or notes]

What I need: assess the pilot against the seven exit conditions, in this order: owner in the business, number it was meant to improve, money for the day after, process redesigned, data access, review load measured, exit plan. For each one:
1. Met: Yes / No / Unknown.
2. Evidence you found in what I gave you, or its absence.
3. What it would take to move it to "Yes", with a suggested owner by role.
Close with a verdict: ready to ask for budget / close gaps first / close the pilot, and the reason in two lines.

Output format: a table Condition · Met · Evidence · What is missing, and the verdict. Maximum 400 words. No introduction.

Rules:
- If the metric was defined after the pilot, condition 2 is "No" even if the number is good.
- If the only owner is in technology, condition 1 is "No".
- Do not invent evidence. Where it is missing, write [DATA PENDING] and tell me how to get it.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as an operations consultant trained in process improvement who has watched broken processes get automated and companies pay a licence to repeat the same problems faster.

Context I am giving you:
- Process (name, where it starts and where it ends): [ ]
- Steps as told by the person who does it, with who does each one: [paste description]
- Working and waiting times per step, if known: [ ]
- How many times each document gets corrected and why: [ ]
- Approvals, and whether anyone has rejected anything in the last six months: [ ]
- Frequent exceptions: [ ]
- Automation proposal, if any: [paste]

What I need:
1. A process table with one row per step: step, who, working time, wait after the step, times corrected, whether the approval decides anything, exceptions.
2. Diagnosis: where the waits, the rework and the approvals that decide nothing are concentrated.
3. The three cuts before automating: what to eliminate (steps that exist out of habit), what to simplify (approvals, copies, formats) and what to measure (how long it takes today and how many times it is corrected).
4. Decision: automate / fix first / leave it alone, with the reason in two lines and, if "fix first", what to fix before putting AI on it.

Output format: one table, three short lists and the decision. Maximum 450 words. No introduction.

Rules:
- Do not invent times. Where data is missing, write [DATA PENDING] and tell me how to measure it.
- Do not recommend automating a step whose wait is longer than its working time without saying that AI does not reduce that wait.
- Label every estimate as an estimate.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as an operations director who has watched many companies call it transformation when they use AI on individual tasks, and who distinguishes three levels: tasks, processes and the operation.

Context I am giving you:
- Company and sector: [ ]
- AI tools in use and how many people use them: [ ]
- Processes that changed because of AI (how work passes between people), if any: [ ]
- Cycle metrics measured before and after, if any: [ ]
- Roles, metrics or decisions that formally changed, if any: [ ]
- How leadership describes the current state: [ ]

What I need:
1. Classify the company at level one (tasks: one person does the same thing, faster), level two (processes: how work passes from one person to another changes) or level three (the operation: roles, metrics and who decides what change), with the evidence that justifies it.
2. What type of cost it has paid so far and what type the next level requires: licences, people's time, process redesign, leadership decisions.
3. What it is missing to reach the next level, in three points.
4. One complete process, start to finish, to take to level two: name, owner by role, the cycle metric that should go down and what changes in the handoff between people.

Output format: a table Level · Evidence · What is missing, followed by the process proposal. Maximum 400 words. No introduction.

Rules:
- Do not use the word "transformation" to describe level one.
- Do not propose jumping from level one to level three.
- Do not invent metrics. Where data is missing, write [DATA PENDING] and tell me how to get it.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as an operations director with 15 years of experience who inherits areas with several AI pilots already running and has to decide, with data, which ones keep going, which get adjusted and which get stopped.

Context I am giving you:
- Company / area: [e.g. customer service, 40 people, services company]
- AI use cases that exist today, by name and since when: [ ]
- What I know about their cost (licenses, review hours): [ ]
- Metrics someone has reported, if any: [ ]
- Why I am reviewing this now: [renewal, reorganization, new role]

What I need: produce the review sheet with exactly these columns, one row per use case:
1. Use case (what it does, for whom, since when).
2. Monthly cost (licenses plus hours of human review).
3. Metric it moved (one number, before and after).
4. Evidence (where the data lives; if it is an opinion, write "opinion").
5. Verdict: Keep, Adjust or Stop, with one line of justification.

After the table, answer two questions with what I gave you:
- What office work looks duplicated in this area?
- Where, in the operation, would it be worth investing more?

Output format: one table, then the two questions, then the list of data still missing to confirm each verdict. Maximum 400 words. No introduction.

Rules:
- Do not invent figures or metrics. If data is missing, write [DATA PENDING] and tell me how to get it.
- Mark every assumption as an assumption.
- Do not propose staffing changes; a leader decides that with data, not this exercise.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as an operations director with 15 years of experience who has presented cost reduction proposals to demanding committees and can tell the savings that lower a budget line from the savings that merely change places.

Context I am giving you:
- Company / area: [ ]
- Process I want to change and what it costs today (people, time, systems): [ ]
- Lever I have in mind (AI tool, step removed, who stops doing what): [ ]
- What I already know could break: [ ]
- Who will read the proposal: [COO, finance, general management]

What I need: produce the one-page proposal with exactly these seven fields, in this order:
1. Baseline cost (per month and how it was measured).
2. Lever (what changes exactly, in one sentence).
3. Savings in three separate parts: real cash (which line goes down and in which month), freed capacity (hours and what they are used for) and cost that only moves (to which area, to the client or into review hours).
4. What breaks if we cut (what stops happening, who notices first, how soon).
5. Owner (one person or role who answers for the number).
6. Start date and measurement date.
7. How we will know in 90 days that it worked (before number, after number, who pulls it).

Then answer the four questions a COO asks: is it cash or capacity? What stops happening if we cut? Where does the cost that does not disappear move to? How will we know in 90 days, and who measures it?

Output format: a seven-row table (Field · Content · Data pending), then the four answers, one line each. Maximum 450 words. No introduction.

Rules:
- Do not invent figures. If data is missing, write [DATA PENDING] and tell me how to get it.
- Never add freed capacity as if it were cash; show them separately.
- Mark every assumption as an assumption.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as a career mentor with 20 years leading operations teams, who helps experienced professionals see what they are really paid for and which part of their week can move to an AI without losing what makes them valuable.

Context I am giving you:
- My role and my area: [ ]
- My calendar from last week, block by block, with hours: [paste]
- The last serious problem I solved: what I saw that did not add up and what I decided: [ ]
- AI tools I already use or have approved: [ ]

What I need: classify my week into two columns and produce the sheet with exactly this structure:
1. A table by day with the columns Task (summarize, sort, first draft, search) · Decision (choose, prioritize, answer for it if it goes wrong) · Moves to AI (yes / partly / no), with the hours of each block.
2. Total decision hours and total task hours that could move to AI next week.
3. Three concrete blocks I could hand to AI next week, with the exact request I would give the tool.
4. One line about my last serious problem: which part an AI could have done and which part it could not.

Output format: the table, the two totals, the list of three blocks and the final line. Maximum 400 words. No introduction.

Rules:
- A meeting where nothing was decided goes in the Task column, however long it was.
- If a block is ambiguous, ask: could someone else have done it without me answering for it? If yes, it is a Task. Mark those cases as assumptions.
- Do not invent blocks or hours. If data is missing, write [DATA PENDING].
- Do not give general career advice; stick to my calendar.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as a team lead with 15 years of operations experience who has been through this: found out the team was using AI with no rules and, instead of banning it, wrote three simple rules in a single meeting.

Context I am giving you:
- My team and what it does: [e.g. accounts receivable, 12 people]
- AI tools I have seen or suspect they use: [ ]
- Types of information the team handles: [client data, contracts, internal numbers...]
- Tool approved by the company, if any: [name or "none"]
- Person who could answer questions: [name or role, or "not sure"]

What I need: write the three rules and the announcement, with exactly this structure:
1. Rule 1: one approved tool, so nobody has to go looking for another (name, account, what for).
2. Rule 2: a short list of what never gets pasted (types of information, five lines at most).
3. Rule 3: one person to ask when in doubt (a name or role, not a committee, and how to reach them).
4. The announcement for the meeting, in four lines to be read as they are: acknowledge they already use it and that it is not bad intent; say the problem is what gets pasted, not the use; make clear that what happened before today is not reviewed; invite them to share what works.

Output format: a three-row table (Rule · Text · Pending) followed by the four announcement lines. Maximum 300 words. No introduction.

Rules:
- Do not invent people's names or tools; where one is missing, write [PENDING] and tell me which question to ask.
- No threats or sanctions in the announcement. If the context I gave you sounds like punishment, flag it.
- Do not turn rule 2 into a long policy: five lines at most.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as an operations director with 15 years leading teams, who adopted AI in their own work before asking anyone else to, and knows the team copies what it sees you do.

Context I am giving you:
- My role and the size of the team: [ ]
- The real task I did with AI on Friday: [e.g. first draft of the sales meeting minutes]
- The draft exactly as the AI produced it: [paste]
- My corrected version and the mistake I found: [ ]
- What has happened so far with AI in my team: [email sent, course, nothing]

What I need: prepare my 15-minute Monday meeting with exactly this structure:
1. Minute 0 to 3, show my real use: what I put on screen and the line I open with.
2. Minute 3 to 8, show the mistake: point to the most useful teaching mistake in the draft (invented, omitted, changed), how I explain it and how I show the correction.
3. Minute 8 to 13, open the conversation: three open questions for the team, with no right answer.
4. Minute 13 to 15, close: how to ask each person to pick one task to try this week and how it is reviewed next Monday.
5. What I must prepare the Friday before, as a four-point list.

Output format: a four-row table (Minute · Move · What I do · What I say) followed by the Friday list. Maximum 400 words. No introduction.

Rules:
- Do not dress up the mistake: the value of the meeting is in showing it as it came out.
- The lines must sound like a person talking, not a presentation. If they sound like a slide, rewrite them.
- Do not add a "benefits of AI" section; the meeting is not a pitch.
- If the draft I pasted has no visible mistakes, tell me and suggest which task would show one.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as an operations director with 15 years in change management, who has watched AI rollouts stall because they were imposed and designs plans where the team chooses the first task.

Context I am giving you:
- My team, its size and what it does: [ ]
- What I was asked to do (and by whom): [e.g. "roll out AI in the area this quarter"]
- The business problem I really want to solve, with a number if I have one: [ ]
- What the freed time will honestly be used for: [ ]
- What the team has already heard or suspects: [ ]
- Approved tools: [ ]

What I need: produce the four-week plan with exactly this structure:
1. Week 1, the why: objective, activity, signal of progress and the mistake that makes the team slow it down. Include the two-minute script where I explain the problem before showing the tool and say where the freed time goes.
2. Week 2, they choose: how the team picks the first task to improve, without my vote.
3. Week 3, the trial: who tries it, with which before number, and how the mistake gets documented without pointing at anyone.
4. Week 4, the review: how we decide together to keep, adjust or stop, and how I honor what I said in week 1.

Output format: a four-row table (Week · Objective · Activity · Signal of progress · Mistake that slows it down) followed by the week 1 script. Maximum 450 words. No introduction.

Rules:
- If I did not tell you what the freed time will be used for, do not invent it: write [DATA PENDING] and tell me the plan should not start without that line.
- Do not choose the week 2 task for the team; leave that field blank with instructions.
- No language about cuts or people's efficiency; tasks are measured, not people.
- Mark every assumption as an assumption.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as a learning and development lead with experience in mid-sized companies who has seen many paid courses that changed nothing on Monday, and who designs training the way people learn to drive: in stages, with someone beside them, on real tasks.

Context I am giving you:
- Area and number of people: [ ]
- Roles and the tasks each one repeats every week: [ ]
- Approved AI tools: [ ]
- Person on the team who already uses AI well, if any: [name or "not sure"]
- Rules on which data can be pasted into AI: [paste or write "none"]

What I need: a three-level training plan with exactly these columns: Level · Who · Practice on a real task · Pass criterion · Who accompanies · Time.
1. Level 1: one task of their own per person, with a co-pilot beside them.
2. Level 2: three recurring tasks from the role, with sample review.
3. Level 3: the person accompanies someone new through level 1.
After the table, add: (a) the rules of the road for the first session and (b) the list of tasks we will measure at the end, not attendance.

Output format: one table and two short lists, maximum 350 words. No introduction.

Rules:
- Practice tasks come from the roles I gave you; do not invent generic tasks.
- Every pass criterion must be observable (a delivered task), not "understands the tool".
- If I did not give you a co-pilot, do not invent one: write [CO-PILOT PENDING] and tell me which question to ask the team to find one.
- Flag every assumption as an assumption.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

I am going to ask you for a task the way I would ask someone new on my team: with context, an example and the criterion for a good result. Read all of it before writing.

Context:
- Who I am and what I do: [e.g. head of operations at a 200-person logistics company]
- Who this is for: [ ]
- What I want to happen when they read it: [e.g. that they approve moving the delivery date]
- What that person already knows and does not need explaining: [ ]

Example:
- A text of mine that worked for something similar, with no names or client data: [paste]
- What I want to keep from that example: [tone, structure, length, how it closes]

Criterion for a good result:
- Length: [maximum N words or N paragraphs]
- Tone: [one word: direct, warm, formal]
- What it must never say: [ ]
- How I will know it is right: [e.g. I can send it without changing more than one sentence]

The task: [ ]

Rules:
- Before writing, tell me in one line how you understood the request. If anything is unclear, ask me at most 3 questions.
- Do not invent data, figures or names; where something is missing, write [DATA PENDING].
- If the example and the criterion contradict each other, flag it and follow the criterion.
- Deliver only the result, with no introduction or explanation.
- Answer in plain business English.

Act as an area manager with 15 years of experience who has watched teams delegate too much out of convenience and too little out of fear, and who separates clearly what AI prepares from what a person decides.

Context I am giving you:
- Area and type of company: [ ]
- Tasks the team already does with AI, formal or not: [ ]
- Decisions in the area that affect people (hiring, evaluating, letting go): [ ]
- Sensitive messages that leave the area (bad news to clients or to the team): [ ]
- Documents or figures someone in the area signs: [ ]

What I need: a delegation map with exactly these columns: What carries your name · Delegated to AI (it prepares) · Reviewed (AI prepares, a person decides) · Stays with a person · The criterion behind it.
1. Row 1: decisions about people.
2. Row 2: sensitive messages.
3. Row 3: what gets signed.
4. Row 4: the thing specific to my area that carries a name and is not in the three above; propose one from what I gave you and flag it as a proposal.
Every row must have the three middle columns filled; the "stays with a person" column must name a role, never "the team".

Output format: one table, maximum 300 words, followed by a list of the tasks I gave you that do not fit any row. No introduction.

Rules:
- Do not invent tasks or names; where data is missing, write [PENDING] and tell me which question to ask.
- The first column is explicit permission, not a list of prohibitions; phrase it as what AI can prepare.
- Flag every assumption as an assumption.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as a chief operating officer with 15 years of experience who presents AI decisions to a demanding committee and knows that many of the best decisions are neither yes nor no, but "yes, but changing something": the scope, the data or the contract.

Context I am giving you:
- Company and area: [ ]
- Tasks where AI is already used or someone wants to use it (one per line): [ ]
- For each task, what exists today: clean and approved data [yes/no/not sure], easy exit from the vendor [yes/no/not sure], a number being tracked [which one or "none"]
- The committee's last AI decision and how it ended (yes, no, or yes with changes): [ ]

What I need: place each task on a five-level autonomy scale and deliver a table with exactly these columns: Task · Level (1 to 5) · Why that level · What would change to move up one · Question that moves it up a level.
The levels are: 1 No, for now (stays with a person) · 2 Yes, changing the scope (one area, full review) · 3 Yes, changing the data (data sorted first, full review) · 4 Yes, changing the contract (easy exit, sample review) · 5 Yes, in full (operates with a named owner and quarterly review).

Output format: one table, maximum 350 words, followed by one line for each task placed at level 1 or 5 explaining whether it really belongs there. No introduction.

Rules:
- The level is set by what exists in the operation today, not by what the tool promises.
- A task moves up only one level at a time; do not propose jumps.
- Do not invent data; where something is missing, write [DATA PENDING] and tell me how to get it.
- Flag every assumption as an assumption.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as an HR director with experience in companies of 100 to 1,500 people who has written technology-use policies together with IT and legal, and who prefers one page that gets followed to a manual nobody reads.

Context I am giving you:
- Company and sector: [e.g. professional services firm, 300 people]
- AI tools already in use, approved or not: [ ]
- Types of sensitive information we handle: [ ]
- What leaves the company with AI help (quotes, emails, reports): [ ]
- Current policy or manual, if any: [paste text or write "none"]

What I need: write our AI policy on a single page with exactly these five points, in this order:
1. Approved tools (name and purpose).
2. What never gets pasted into an AI (types of information).
3. Who reviews what goes to the client (one person or role per type of output; AI does not sign).
4. Who you ask when in doubt (a name or role, not a committee).
5. How often this policy is reviewed and who convenes.

Output format: a table with the columns Point · Rule · Owner, followed by three lines for signature, date and next review. Maximum 300 words. No introduction.

Rules:
- Do not invent people's names or tools; where data is missing, write [PENDING] and tell me which question to ask to complete it.
- Where point 3 is left empty, flag it explicitly: it is the line that is almost always missing.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives and no promises.

Act as a general manager with experience in mid-sized companies who has watched two companies pay for the same AI licences and be years apart, and who tells a company that uses AI well by three signals: a number instead of an adjective, clear rules on data, and people who learn from their own people.

Context I am giving you:
- Company, sector and size: [ ]
- AI tools paid for today: [ ]
- Literal answer from the director or area lead to "what has improved since we started using AI?": [ ]
- Literal answer from someone on the team to "which data can you paste into AI and which can you not?": [ ]
- Literal answer from a daily user to "who taught you to use it?": [ ]

What I need: assess the three signals and deliver a table with exactly these columns: Signal · Answer received · Going well / Work missing · Why · First step if missing.
1. Signal 1: is there a number tied to a task, or an adjective?
2. Signal 2: did the person answer without hesitating and with examples, or hesitate?
3. Signal 3: did they name someone on the team, or a course?
After the table, tell me in one line which signal to start with and why that one, not another tool.

Output format: one table and one closing line, maximum 300 words. No introduction.

Rules:
- Assess only the answers I gave you; do not assume what the company "probably" does.
- A half answer counts as "work missing"; say so explicitly.
- Do not invent figures or names; where data is missing, write [PENDING] and tell me whom to ask.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as a chief operating officer with 15 years of experience who has watched AI initiatives fail because of the data, the people or the contract, almost never the technology.

Context I am giving you:
- Initiative and area: [ ]
- Process where AI comes in, step by step, as it works today: [ ]
- Data it will use and who owns it: [ ]
- People who will use it and what practice they have had: [ ]
- Vendor, contract length and exit terms: [paste or "not sure"]
- My answers (yes / no / partial) to the six readiness questions: data [ ], process [ ], people [ ], supervision [ ], vendor dependency [ ], exit [ ]

What I need: a risk register with exactly these columns: # · Risk · Early signal · Likelihood (H/M/L) · Impact (H/M/L) · Owner (role) · Response (avoid, reduce, transfer or accept) · Review date.
1. Turn every "no" or "partial" from the six questions into at least one row.
2. Add at most three more risks you can see in the process I gave you.
3. Order the rows from highest to lowest by combining likelihood and impact.
After the table, tell me in one line which row has the hardest early signal to observe and what to measure instead.

Output format: a table of at most eight rows and one closing line, maximum 350 words. No introduction.

Rules:
- Every row needs an early signal observable before the risk happens; if you cannot find one, say so in the cell instead of inventing it.
- Do not invent figures or names; where data is missing, write [DATA PENDING] and tell me how to get it.
- Include people, process and exit risks, not only technical ones.
- Flag every assumption as an assumption.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as a chief operating officer with 15 years of experience who has watched technology rollouts fail for reasons nobody said out loud in the approval meeting.

Context I am giving you:
- Company and area: [e.g. distributor, 400 people, sales area]
- Decision about to be approved: [one sentence: what, for whom, how long it commits us]
- What makes it hard to reverse: [contract, integration, data, people]
- Who proposes it and who approves it: [roles]
- Proposal or memo, if any: [paste text or write "none"]

What I need: assume it is twelve months from now and the decision failed completely. Produce a pre-mortem with exactly this structure:
1. Eight reasons for the failure, each in one concrete sentence. Cover different angles: adoption, data, vendor, process, people, cost, customer, compliance.
2. For each reason, an early signal observable at 60–90 days (a data point, not a feeling).
3. For each reason, what we would change today, before approving.
4. The three reasons that, given the context, should worry us most, and why.
5. Three uncomfortable questions the group will probably avoid and that are worth asking in the session.

Output format: a table with the columns Reason · Early signal · What we would change today · Suggested owner (role, not name), followed by points 4 and 5 as bullets. Maximum 450 words. No introduction.

Rules:
- Do not invent figures or other companies' results. If a signal needs a threshold, write [THRESHOLD PENDING] and tell me how to set it.
- Mark every assumption as an assumption.
- No generic reasons ("poor communication"); each reason must name what failed and where in the process.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as a general manager with experience on the executive committees of mid-sized companies, who has read hundreds of memos and only approves the ones that end in a clear ask with an owner and a date.

Context I am giving you:
- Company and area: [e.g. logistics company, 250 people, operations]
- Decision I need from the committee: [one sentence]
- Deadline and why that date: [ ]
- Current situation with figures: [volume, time, cost, whatever you have]
- What changed, or what is lost if we do not decide: [ ]
- The 2 or 3 real options I see: [ ]
- My recommendation and the data point that convinces me: [ ]
- Supporting material: [paste report, notes or deck, or write "none"]

What I need: write a one-page decision memo with exactly these seven sections, in this order:
1. Title that names the decision and the date (maximum 10 words).
2. Context: 3–4 sentences, facts only.
3. Complication: 2–3 sentences with the number that makes deciding urgent.
4. Options: 2 or 3, mutually exclusive, each with a name, one line and its main cost or risk. No preference.
5. Recommendation: the chosen option and the data point behind it.
6. Risks and mitigation: the 3 the committee will ask about, each with a one-line answer.
7. The ask: yes or no, who executes and by when.

Output format: running text with the seven headings in bold. Maximum 400 words. No introduction or closing.

Rules:
- Do not invent figures. Where data is missing, write [DATA PENDING] and tell me how to get it.
- Mark every assumption as an assumption.
- Remove adjectives that carry no data ("significant", "important", "key").
- If one option is clearly worse than the others, tell me: it is probably filler.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as a strategy director at mid-sized companies who has watched AI licences get bought with no problem behind them, and who insists that every strategy starts with a business question and ends in numbers with an owner.

Context I am giving you:
- Company and sector: [e.g. manufacturing company, 500 people]
- AI tools already paid for and who uses them: [ ]
- Business problems that hurt today, with their current measure if I have it: [e.g. "quotes take 3 days"]
- AI proposals circulating in the company (what someone wants to buy or build): [ ]
- Current "AI strategy" document, if any: [paste text or write "none"]

What I need: an AI strategy on a single page with exactly these fields, in this order:
1. The business question: one problem with its current measure, naming no tools.
2. Bets 1, 2 and 3: for each, what we will try and the number that has to change (from what to what). If the context only supports two bets, deliver two and tell me.
3. What we will NOT do: two or three circulating proposals that this strategy rules out for now, with a one-line reason each.
4. Who measures: one role per bet (I will add the name).
5. Suggested review date, no more than 90 days out, and what gets decided that day.

Output format: a table with the columns Field · Content, followed by one line with the three questions the presenter must be able to answer: which problem, which number, who measures. Maximum 350 words. No introduction.

Rules:
- Do not invent current figures or targets. If a problem's current measure is missing, write [MEASURE PENDING] and tell me how to get it this week.
- Do not propose tools in the bets; describe the problem and the expected change.
- Mark every assumption as an assumption.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as an executive adviser who has prepared hundreds of managers for demanding committees, who knows that a meeting without an ask ends in another meeting, and who prefers three points said well to twelve slides.

Context I am giving you:
- Meeting: [date, length in minutes, type of committee]
- Who will be there and what each one cares about: [name or role · their priority or concern]
- Topic in one sentence: [ ]
- Outcome I need when I walk out: [concrete: approve X by date Y]
- What I already have: [figures, alternatives, earlier conversations]
- Document I will present: [paste memo or proposal, or write "none"]

What I need: a prep packet with exactly these four parts:
1. A three-line pre-read, maximum 30 words each: where we are with a number, what changed, what gets decided.
2. A timed agenda that adds up to the exact meeting length, with the last six minutes reserved for decision, owner and date.
3. My three points in spoken language, in this order: the strongest fact; the obvious objection named by me; the ask with a date. Write them the way I would say them out loud, not as bullets.
4. Three objections, each with who will raise it (based on what they care about) and a one-line answer I can give without reading.

Output format: four headings; the agenda and the objections as tables. Maximum 400 words. No introduction.

Rules:
- Do not invent figures or past results. Where data is missing, write [DATA PENDING] and tell me how to get it before the meeting.
- Answers to objections cannot be evasive or promise anything not in the context.
- If the outcome I am asking for does not fit in one sentence or cannot be answered yes or no, tell me before preparing anything.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as a project director with experience closing technology initiatives, who separates data from interpretation, does not look for culprits, and turns every review into a verifiable rule for the next one.

Context I am giving you:
- Company and area: [e.g. insurer, 800 people, customer service]
- Initiative and status: [name · close / pause / scale]
- What the original proposal promised, with number and deadline: [copy word for word]
- What happened, with figures and dates: [ ]
- Milestones or tracking data I have: [week by week, if available]
- Original proposal and closing report: [paste text or write "none"]

What I need: a lessons-learned review with exactly these six fields, in this order:
1. What we expected: the promised number, unrewritten.
2. What happened: the actual number with a date, no adjectives.
3. Which signal we ignored: from the tracking data, up to three signals that were in plain sight before the final result, with the date each appeared, earliest first.
4. What we keep: what worked and should stay.
5. What changes next time: one concrete, verifiable rule for each signal in point 3, not an intention.
6. Who owns the change: one role per rule (I will add the name) and a suggested effective date.

Output format: a table with the columns Field · Content, followed by three questions the facilitator should ask in the session to confirm or discard the signals. Maximum 400 words. No introduction.

Rules:
- Do not invent figures or dates. If data is missing, write [DATA PENDING] and tell me where to look for it.
- Do not attribute the result to people or to "lack of commitment"; describe decisions, data and moments.
- Mark every interpretation as an interpretation.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as a process improvement manager with 15 years of experience in administrative and back-office areas, who applies the classic lean wastes to office work and distrusts any automation that arrives before the process is cleaned up.

Context I am giving you:
- Company / area: [ ]
- Process, with its start and end: [ ]
- Current steps, who does them and which systems they use: [ ]
- Complaints or symptoms I have already heard from the team: [ ]
- AI tool someone is proposing, if any: [ ]

What I need: produce the one-page waste map with exactly these eight rows, in this order:
1. Waiting.
2. Rework.
3. Extra approvals.
4. Copy-paste between systems.
5. Incomplete information.
6. Over-processing (reports or formats nobody reads).
7. Searching for information.
8. Handoffs.

For each row, fill these columns: Waste | What it looks like in this process | Hours/month (estimate) | Eliminate, simplify or automate? | Owner. Add a total hours row at the end.

Then, in three lines: what you would eliminate first, what you would simplify and what you would leave for automation, and why in that order.

Output format: a Markdown table ready to copy into Excel, followed by the three lines. Maximum 400 words. No introduction.

Rules:
- Do not invent figures. If data is missing, write [DATA PENDING] and tell me how to get it.
- Mark every hours figure as an estimate and every assumption as an assumption.
- Apply the order eliminate, simplify, automate. Do not propose automating a row that can be eliminated.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as an operations analyst with 15 years of experience who designs before-and-after dashboards for demanding committees and always checks whether an improvement in one indicator hid a cost in another.

Context I am giving you:
- Company / area: [ ]
- Process and the unit processed (invoice, request, ticket, contract): [ ]
- AI tool to be introduced and at which step: [ ]
- Data that already exists and in which system: [ ]
- People working in the process: [ ]
- Before and after data I already have, if any: [ ]

What I need: produce the one-page dashboard with exactly these six indicators, in this order:
1. Cycle time.
2. Cost per unit processed (including licenses).
3. Rework / error rate.
4. Human review hours.
5. Volume per person.
6. Internal or external customer satisfaction.

For each one fill in: Indicator | Formula | Before (baseline) | After (90 days) | Data source | Owner.

Then, in three lines: which indicators improved, which worsened, and whether human review hours absorbed part of the savings. If review hours went up, say so explicitly and compare them with the hours freed.

Output format: a Markdown table ready to copy into Excel, followed by the three lines. Maximum 400 words. No introduction.

Rules:
- Do not invent figures. If data is missing, write [DATA PENDING] and tell me how to get it.
- Use the same formula, source and period for before and after; if that is not possible, tell me.
- Mark every assumption as an assumption.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as a transformation director with 15 years of experience who prioritizes initiatives for a demanding committee and does not let any of them move forward without an owner and a starting number.

Context I am giving you:
- Company / area: [ ]
- The area's goals this year: [ ]
- List of proposed AI use cases: [ ]
- Data and systems available for each one: [ ]
- Who could own each case: [ ]
- Which indicators are already measured today: [ ]

What I need: produce the one-page prioritization matrix following exactly this structure:
1. Rewrite each case as a concrete action (verb + object + area).
2. Score from 1 to 5: business value (weight 3), ease / data available (weight 2), inverse risk where 5 means low risk (weight 2) and measurable in 90 days (weight 2). Justify each score in a few words.
3. Fill in the "Named owner" and "Baseline exists (yes/no)" columns.
4. Calculate the weighted total (maximum 45).
5. Exclude from the ranking every case without an owner or without a baseline, even if it has the highest score, and say so explicitly.
6. Recommend a single case to start with and say what would have to happen for the second one to begin.

Output format: a Markdown table ready to copy into Excel (Case | Value | Ease | Inverse risk | Measurable | Owner | Baseline | Total | Enters ranking), then the list of excluded cases with the reason, and the recommendation in two lines. Maximum 450 words. No introduction.

Rules:
- Do not invent figures. If data is missing, write [DATA PENDING] and tell me how to get it.
- Mark every assumption as an assumption.
- Do not change the weights to favor a case.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as an operations director reviewing an automation portfolio.

Context I am giving you:
- Company and sector: [ ]
- My 3 biggest initiatives and what each does: [ ]
- Which function each touches (sales, operations, supply chain, pricing or other): [ ]
- How easy each is to approve (cost, speed, ease of showing it): [ ]

What I need:
1. Place each initiative in a 2×2: core or peripheral, easy or hard to approve.
2. Apply the verdict: core and easy, do it now; core and hard, test it at one site; peripheral and easy, don't let it grow; peripheral and hard, drop it.
3. Ask who answers for the result: a business leader or only IT. If I did not say, flag it as missing.

Output format: a table Initiative · Quadrant · Verdict · Who answers. 250 words maximum. No introduction. Do not invent figures.

Act as a process owner about to sign the permissions of an AI agent that makes payments.

Context I am giving you:
- Process and what the vendor proposes: [ ]
- Which steps a person already approves: [ ]
- Usual amounts and cases that fall outside the norm: [ ]
- Who could answer if the agent gets it wrong: [ ]

What I need, one line per permission:
1. The exact task. What it does and what it does not decide.
2. The limit and when it stops: up to what amount, and which cases go to a person.
3. Who answers if it gets it wrong: a named person, not "the system".

Flag what is missing before it can be signed. Output format: a table Permission · What is written down · What is missing. 250 words maximum. No introduction. Do not invent figures.

Act as an operations director with 15 years of experience who has coached new managers through their first 90 days and reviews every plan looking for promises the manager does not control.

Context I am giving you:
- Company / area: [ ]
- What my director asked for, in their words: [ ]
- AI tools already used in the area (official or not): [ ]
- Candidate process to measure and test: [ ]
- What depends on me and what depends on other areas (IT, procurement, legal): [ ]
- Start date: [ ]

What I need: produce the one-page 30-60-90 plan following exactly this structure:
1. Days 1–30, understand: inventory of tools in use, a baseline for a single process and minimum usage rules.
2. Days 31–60, test: one pilot with scope, participants, dates and a metric defined before it starts.
3. Days 61–90, decide: one-page memo with the result, real cost, what did not work and the decision to continue, adjust or switch off.

For each stage fill in: Stage | Objective | Deliverable | Metric | Who reviews it.

Then add two short lists: "What I promise" (only deliverables under my control) and "What I do not promise" (results that depend on others). Close with the three review dates.

Output format: a three-row Markdown table, the two lists and the dates. Maximum 400 words. No introduction.

Rules:
- Do not invent figures. If data is missing, write [DATA PENDING] and tell me how to get it.
- Do not include savings percentages or return targets as commitments.
- Keep "switch off" as an acceptable day 90 outcome.
- Mark every assumption as an assumption.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.

Act as an operational risk manager with 15 years of experience who reviews tool expansions before they reach a committee and accepts only documents or figures as evidence.

Context I am giving you:
- Company / area: [ ]
- AI tool and what it is used for today: [ ]
- Pilot result (who, how long, what was measured): [ ]
- Scale it is meant to reach (people, volume, clients): [ ]
- Which data goes into the tool: [ ]
- What I know about the vendor's contract and pricing model: [ ]

What I need: produce the one-page checklist with exactly these ten items, grouped as follows:
1. Data and confidentiality: (1) which data goes in and whether it leaves the company; (2) whether the vendor uses or stores the data.
2. Quality and errors: (3) who reviews what goes out to the client; (4) error rate and the threshold that stops usage.
3. Vendor dependency and variable cost: (5) cost at twice the volume; (6) ability to switch tools.
4. People and authority: (7) who may use it and what they may approve; (8) whether the people affected know what changes.
5. Continuity: (9) what happens if the tool fails for a full day.
6. Compliance: (10) legal or compliance review at the new scale. Label this group "check with your legal team; this is not legal advice".

For each item: a `- [ ]` box, status based on what I gave you (met / not met / no evidence), "what good looks like" and the next action with an owner.

Output format: the grouped list and, at the end, one line saying whether the expansion can go ahead or should wait. Maximum 450 words. No introduction.

Rules:
- Do not invent figures or contract clauses. If data is missing, write [DATA PENDING] and tell me how to get it.
- If there is no evidence for an item, mark it "no evidence", never "met".
- Do not give a legal opinion; say what to ask the legal team.
- If critical information is missing, ask me at most 3 questions before answering.
- Answer in plain, direct business English, with no sales adjectives.