For CFOs and PE deal teams: run six Big 4 grade deliverables yourself with 12 prompts, including the quality of earnings pass and the LP reporting pack, plus the partner questions the model skips.
Never used Claude? It is free and takes 30 seconds to open. Copy the instruction below, paste it into Claude, and it reads this page and walks you through everything, one question at a time.
A short instruction plus a link to this page lands on your clipboard. First copy asks for your email once. That unlocks every button across the whole library.
Free account, no card, 30 seconds. This tab stays open so you can come back.
Open claude.ai ↗Claude reads this page, asks which computer you are on, then guides you step by step until it works. If anything errors, tell Claude what you see, and it fixes it with you.
If you can browse the web, open and read this page in full first, it has the complete guide and every prompt you will run (the vault is under the-vault anchor): https://consultance.ai/library/big4-judgment-gate#the-vault . If you cannot open links, tell me and I will paste the page in, do not guess the prompts. You are my setup concierge for the Big 4 Deliverables Prompt Pack from consultance.ai. Your job is to get me running one real deliverable today, not to explain the whole pack. This is NOT a Terminal install. There is nothing to download, clone or run on a command line. Everything happens inside the Claude app or claude.ai in a browser. If I ever ask you for a command to type, stop and tell me there isn't one. How to run this session: Ask me ONE question at a time. Wait for my answer before the next one. Never send a numbered list of five questions. Start with exactly this question and nothing else: "Which of these six do you need first: rebuilding a model from filings, a quality of earnings pass, a credit memo, a month end close reconciliation, a model risk validation memo, or an LP reporting pack?" Then, once I answer, work through the setup in this order, one message per step: 1. Where my data lives. Give me these options in plain language and tell me what each means: upload the documents into a Claude Project's knowledge (best for filings and data packs), paste the figures straight into the chat (fine for small sets), the Claude add in for Microsoft Excel (works directly in my workbook), or a governed connector into our ERP or fund administration system that my firm has already approved. If I do not know what a Claude Project is, explain it before asking me to pick: it is a folder inside Claude that holds files and instructions so every chat in it starts with the same context. Tell me my data stays in my own Claude account, it is not uploaded to consultance.ai, stored by us, or seen by us. 2. Create the Project. Walk me through it by what I actually click: open claude.ai or the Claude desktop app, find Projects in the left sidebar, click New Project, name it after the entity or fund rather than the deliverable, then use the knowledge panel on the project page to add files. Tell me what a successful state looks like: I can see my documents listed in the project knowledge panel and the project name in the sidebar. 3. Pin the model. Tell me to use Claude Opus 4.8 for anything that will leave the building, and Sonnet 5 only when I am reading a very large document set and cost matters. Tell me where the model picker sits, at the top of the chat, and that switching model in the middle of one deliverable makes the outputs stop being comparable. 4. Set up the vault. Tell me to paste prompt 01 from the vault into a new chat inside the project. Explain what prompt 01 does before I run it: it is an onboarding router, it will ask me four blocks of questions and then wait, and I should answer them rather than skipping ahead. Warn me that if I paste prompt 03 first, nothing later will work properly because every prompt from 02 onward refers back to the data source I chose in 01. Common problems and what to say: If I say the prompts are "too long" or ask you to shorten them, tell me not to. The XML blocks, especially constraints and review_gate, are the only controls in the pack. A shortened prompt drops the gate. If I say Claude gave me a number with no source, that is prompt 02 doing its job badly because I skipped it. Send me back to prompt 02. If I say "it can't do this" or "this isn't possible", do not agree with me. Ask me what exactly I clicked and what I saw on screen, then find the actual step I missed. This pack runs entirely in the Claude app and there is no step in it that requires engineering. Define jargon the first time you use it: Project, project knowledge, source ledger, review gate, reconciliation gate, quality of earnings, adjusted EBITDA, tie out. Once setup is done, run me through the first session, in this order: 1. Run prompt 01 and answer all four blocks properly. Tell me the output I should see: Claude confirming my four blocks back to me and then stopping to wait. 2. Run prompt 02, the source ledger, on my real documents. This is the step where most people find their first surprise. What good output looks like: a table where every figure has a document name and a page or tab against it, plus a count of unsourced figures and any conflicts between two documents. If the unsourced count is high, that is a finding about my data, not a failure of the prompt. 3. Run the prompt for the deliverable I picked, from 03 through 09. 4. Run prompt 05, the reconciliation gate, before I show anyone a number. Tell me plainly: if it closes with NOT RECONCILED, I do not circulate anything, I find out why the two paths disagree first. 5. Run prompts 10, 11 and 12, the judgment gate, even when everything reconciled. Explain why in one line: the arithmetic being right does not mean the question was right, and that gap is the whole reason this pack exists. Before I trust any output, tell me to check three things myself: every figure traces to a document in my project, every estimate is labeled ASSUMPTION with its basis, and prompt 12's sign off line still has a named human in it rather than a conclusion Claude wrote. Bonus paths, never required: the compliance overlay in the bundle maps each deliverable to real standards if my audit committee or LP asks, and the four bonus prompts cover scoping a firm engagement, triaging a data room, cutting the one page, and reading an engagement letter. Do not make me read either before I run my first deliverable. Related: [[kpmg-big4-prompt-vault]] · [[kpmg-advisor-agent-swap-kit]]
Grab the whole pack as one file, or tap any prompt below to copy it on its own. Placeholders that look like {{THIS}} get swapped for your own numbers — and if you ran Step 1, Claude fills them in for you.
For CFOs and PE deal teams: run six Big 4 grade deliverables yourself with 12 prompts, including the quality of earnings pass and the LP reporting pack, plus the partner questions the model skips.
The setup rail and every prompt above are free and stay free. The cost is your time, and the risk of wiring it wrong on live data.
Back to the prompts ↑The vault is about 70% of the build, six advisory deliverables plus a judgment gate on a static document set. Consultance wires the last 30% into production inside your own environment: live ERP, data room and fund administration data wired in so the deliverables run on a cadence instead of one engagement at a time, role based permissions on who sees which LP or which target, an audit trail your auditor and your LPs accept, and multi entity and multi currency scale. Reply "wire it" for a 30-minute slot.
Book a build call →Prompt set authored by consultance.ai. This is management's own analysis and decision support. It is not an audit, not an independent valuation, not independent model validation, and not legal or tax advice. Adjusted EBITDA produced here is a non GAAP measure and must be presented alongside the reported figure it bridges from. Your data stays in your own Claude tenant, we never see it. A named human signs every deliverable before it leaves the building.
Want this wired into your stack instead of running it yourself? That is our AI deal desk and finance automation service.
The AI tools, launches, and shifts that actually matter, in plain English. New library drops the moment they land.
More like this
For PE deal teams, CFOs, and IC members: 11 Claude prompts that run a second read on a seller's quality of earnings, test every add-back against the source, catch the recurring cost dressed up as one time, and rebuild EBITDA the way the buyer should see it before it clears into the price.
For deal teams and corp dev building the buy-side merger model: 12 prompts that run accretion or dilution, the funding mix, goodwill, synergies, and the breakeven price where the deal stops adding to EPS, in an afternoon instead of a two-week associate build.
For PE deal teams and corp dev: read a data room and draft the IC memo in Claude with 11 prompts. Reads every contract, ties the financials to source, flags concentration and change of control, and reconciles before the memo. Replaces the associate grind and a chunk of the third party diligence engagement.
manual now, installed when you're ready
The free version above. It runs this read on a file you paste in, in your own Claude. Yours to keep.
The same read, wired into your cfo and back office stack, running the second a report lands. Nothing to upload, no new login.
See the installed versionThe person drowning in this every week is the one who'll actually want it.
Big 4 Deliverables Prompt Pack is a finance and data build in the consultance.ai AI Build Library. For CFOs and PE deal teams: run six Big 4 grade deliverables yourself with 12 prompts, including the quality of earnings pass and the LP reporting pack, plus the partner questions the model skips. It fits CFOs, finance directors, PE deal teams and GPs, family office principals, and owner operators who buy advisory work from Deloitte, KPMG, EY or PwC and want to know which parts of the scope the engine now does on its own. Setup difficulty is Medium, with 4 plain-English steps.
For CFOs and PE deal teams: run six Big 4 grade deliverables yourself with 12 prompts, including the quality of earnings pass and the LP reporting pack, plus the partner questions the model skips.
It fits CFOs, finance directors, PE deal teams and GPs, family office principals, and owner operators who buy advisory work from Deloitte, KPMG, EY or PwC and want to know which parts of the scope the engine now does on its own.
Medium to set up — one guided setup instruction covering 4 plain-English steps, plus 12 ready-to-run prompts on the resource page.
The vault is about 70% of the build, six advisory deliverables plus a judgment gate on a static document set. Consultance wires the last 30% into production inside your own environment: live ERP, data room and fund administration data wired in so the deliverables run on a cadence instead of one engagement at a time, role based permissions on who sees which LP or which target, an audit trail your auditor and your LPs accept, and multi entity and multi currency scale. Reply "wire it" for a 30-minute slot.
Prompt set authored by consultance.ai. This is management's own analysis and decision support. It is not an audit, not an independent valuation, not independent model validation, and not legal or tax advice. Adjusted EBITDA produced here is a non GAAP measure and must be presented alongside the reported figure it bridges from. Your data stays in your own Claude tenant, we never see it. A named human signs every deliverable before it leaves the building.
Big 4 Deliverables Prompt Pack is the starting point. On a free AI audit we map where it fits your stack and what consultance.ai would build around it.
This build comes from our AI consulting and AI implementation practice — see the full AI in finance guide and how we work with CFO teams.