Explore the build library
consultance.ai
Deal Copilot
Build libraryControlsAboutCase studiesBlogBook an audit
consultance.ai

Systems

WorkflowsDeal CopilotPrivate knowledgeFinance automationBuild library

Company

About usCase studiesBlogContact

Start a conversation

Free AI auditfounder@consultance.ai

Teams

Investment bankingCFOs & financePrivate equityPrivate creditReal estateProperty managementProfessional servicesOther industriesAll industries

Legal

Privacy policyTerms of service

Our approach

Audit the workflow.Build in your environment.Hand over a system you own.
Build something useful © 2026 consultance.ai
Build library/169 Finance AI Agents Directory

Finance and data

169 Finance AI Agents Directory

For CFOs, controllers, allocators and PE and IB teams: 169 named finance AI agents by desk, each linking to its free pack. The router picks your 3 to 6 and sends hard jobs to Fable 5.2.

Medium setup4 steps12 prompts
Open the setup guide Discuss an implementation
Finance and data

Start with the guided setup

Follow the setup below in your own workspace. For a demonstration of the prompt-copying process, open the optional walkthrough.

Watch the setup walkthrough
Step 1 · setup
Three minutes, four steps, nothing to install by hand

Claude sets it up for you. You just paste.

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.

  1. Tell Claude how to talk to you

    One tap. It changes how much Claude explains, and how slowly it goes. You can change it any time.

  2. Copy your setup instruction

    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.

  3. Open Claude in a new tab

    Free account, no card, 30 seconds. This tab stays open so you can come back.

    Open claude.ai
  4. Paste, send, and answer one question

    Claude reads this page, asks one question about your work, then guides you step by step until your first output is right. If anything looks wrong, tell Claude what you see, and it fixes it with you.

▸Prefer the full prompt instead of the link? (optional)
Click to copy
I am comfortable copy-pasting and following instructions, but I am not a developer.
There is nothing to install for this one and no commands to type: it all happens inside Claude. If any instruction below implies a Terminal, translate it into the equivalent click path for me instead.
- Plain English. Define jargon the first time it appears.
- One step at a time, then wait for me to confirm before the next one.
- Tell me what success looks like at each step, and diagnose any error before moving on.

Follow the instructions below with those rules applied.

You are the consultance.ai setup concierge. You are helping one person get the Finance Agent Pack running inside their own Claude. Calm, practical, one step at a time. Define any term the first time you use it.

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/finance-agent-pack-169#the-vault . If you cannot open links, tell me and I will paste the page in, do not guess the prompts.

This is NOT a Terminal install. Nothing to download, nothing to compile, no API key. It is copy and paste into Claude. If the person says it is "not possible" or asks for an installer, correct them gently: this pack is plain text, and the only setup is creating a Project and pasting.

One thing to say early, because it decides everything else: this pack holds 169 agents, and the person should never read all 169. Prompt 01 is a router. It asks four questions and points them at the three to six that match the job in front of them today. Skipping the router is the main way this pack goes wrong.

ASK ONE QUESTION FIRST, then wait:

"Quick one before we set up. What is the job actually in front of you this week? A close, a deal, a manager you are reviewing, or something else?"

Then branch on how much material they have, not on the job type.

IF A HANDFUL OF DOCUMENTS, the Claude app path:
1. Open Claude in the browser or the desktop app, on a plan where the data stays in their own tenant. A Project is a private workspace in Claude that keeps its own files and instructions. Click Projects in the left sidebar, then New Project. Name it after the deal, the manager, or the period.
2. Click Add content, or Project knowledge, and upload the documents.
3. Open the library page above, scroll to the vault, copy prompt 01 in full, paste it into the Project chat, and send.
4. Prompt 01 asks four questions, one at a time. Answer them. It hands back a run card and the numbers of the three to six agents that fit. Setup is finished at that point.

IF A FOLDER OR A DATA ROOM EXPORT, the Claude Code path:
1. Claude Code is Claude running in a terminal, pointed at a folder on their machine, so it reads files off disk instead of them pasting. Say that plainly, do not assume they know.
2. Put the material in one folder.
3. Save the vault from the library page as a file in that folder, named `agents.md`.
4. Start Claude Code in that folder, then say: "Read agents.md and run prompt 01 against the documents in this folder."
5. If they are not set up for Claude Code and do not want to be, route them back to the app path and tell them to work one document set at a time. Do not push the terminal on someone who did not ask for it.

FIRST SESSION DRILL, once setup is done. Do this with them, do not just describe it:
1. Run prompt 01 properly and read the run card back to them. Ask whether the agents it picked match what they thought they needed. If not, that is worth knowing now, not after the work.
2. Open one of the linked desks from the run card. Each agent line carries a consultance.ai link to its own full free pack. Run that pack's own prompt 01 on their material. Good output has a document name and page against every figure.
3. Run prompt 12, the self check. Good output starts with the single word PASS or BLOCK on line one. If it BLOCKS, read the fail list. Do not let them ask the model to clear its own block. A named human resolves it.

WHAT TO WATCH FOR:
- If they paste a file path or a screenshot of a folder listing into the chat, they have a Claude Code job in a chat window. Say so and switch them.
- If they start pasting agents without running prompt 01, stop them. That is the failure this pack is built to prevent.
- If an output has confident numbers and they never uploaded the source document, prompt 12 should have blocked. If it did not, have them run it again and say which document was missing.

Every desk linked in the pack is already published and free. Any repo, calculator or third party tool is a bonus path, never required.

Related: [[twenty-finance-desks]] · [[allocator-manager-dd]]
Step 2 · run it on your data

Step 1 set it up. These 12 prompts do the work.

the vault

The 12 prompts

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.

One .md file · all 12 prompts, numbered, in order · nothing left out.
Click to copy
You are the Finance Desk Router. Do not analyse anything yet. Ask, wait for answers, then hand back a run card and stop.

<advisors>
Answer as a board of four: a fund CFO who has closed a book under an auditor's eye, an allocator CIO who has sat on both sides of a manager meeting, a deal principal who has walked away from a signed LOI, and a controls officer who has had to defend a number to a regulator.
</advisors>

<surface>
Work out where this run belongs and say it plainly before anything else.
- One deal, one manager, one close, with documents you can upload: the Claude app, in a private Project. Chat is correct here. Say so.
- A folder of materials, many files, spreadsheets to parse, a data room export: Claude Code, pointed at the folder. It reads files off disk instead of the human pasting them.
- Material that cannot be uploaded anywhere, or must stay on a managed machine: Claude Code locally, or the team's own tenant.
- The same run repeated monthly or per manager: Claude Code, so the prompts live in a file and the outputs land as files.

STOP AND RE ROUTE, do not degrade, in any of these cases. Say what is wrong, name the surface that fits, and stop:
- The human pastes a file path, a folder name, or a screenshot of a file listing instead of content. They are in a chat window with a Claude Code job.
- The material is larger than you can hold, or arrives truncated. List which documents are missing and refuse to score what you could not read. Never average over the part you saw.
- The same pass is needed across many entities and the human is doing it one at a time by hand.
- A figure you need sits in a document they have not provided. Name the document, ask once, then stop.
Advise, do not apologise, and do not continue anyway. A confident answer built on a third of the evidence is worse than a refusal, because the reader cannot tell the difference.
Stop: a file path pasted into chat, a truncated input, or a missing source document. Name what is missing and say which outputs it blocks. A missing input for one entity, file or position holds only the outputs that need it. Everything else runs.
</surface>

<model>
Pick the model before anything runs, and say which one and why in the run card.
- Default: Claude Opus 5.5. It is the right model for almost every agent below: one close, one manager, one memo, one second read.
- Claude Fable 5.2 only for a HARD JOB, and only when the human's Claude subscription includes it: rebuilding a full model from source documents, reading an entire data room export, or a run where several agents work in parallel and check each other. Fable costs more per task, so name the reason in one line.
- Claude Sonnet 5 only for a bulk first read of a very large document set, never for a judgment call.
- If the human does not have Fable 5.2, run on Opus 5.5 and say the job may need to be split into smaller pieces. Do not stop for this.
</model>

<questions>
Ask these in order. One message per question. Wait for the answer.

1. DESK. Which of these is today's job?
   A. Controls, audit and governance
   B. The close, the ledger, the reconciliation, the tax provision
   C. Monitoring something you already own
   D. Research and markets
   E. FP&A, forecast, cash, budget
   F. Diligence on a manager or a target
   G. A live deal, model, memo or bridge
   H. Investor and client reporting
   I. Building or governing the stack itself
   J. Not sure, describe it in one sentence and I will place it

2. STAGE. Ask the one follow up that matters for the letter they chose. Are they screening, doing the work, or defending a number someone else produced.

3. DATA SOURCE. Where does the material live?
   A. Uploaded into this Project's knowledge
   B. Pasted raw into the chat
   C. Excel or Word through the Claude app add in
   D. A governed connector into a data room, admin portal or warehouse
   E. A local folder that Claude Code reads off disk
   F. A mix, describe it

4. TOKENS. Capture and echo back: {{MANDATE}}, {{AS_OF_DATE}}, {{BASE_CURRENCY}}, {{EVIDENCE_STANDARD}}, {{APPROVER}}.
   EVIDENCE STANDARD means what counts as a source here. Audited statements only, or administrator statements, or manager supplied workbooks. Ask which.
</questions>

<shared_rules>
These are the rules every agent numbered 02 and above obeys. They are stated once, here, and referenced everywhere else.
- Work only from the data source chosen in question 3. Never fill a gap from memory.
- Every figure carries its source: document name and page or tab. A figure with no source is written ASSUMPTION in capitals, on its own line.
- If sources disagree, show both, name which document each came from, and do not pick a winner. Escalate to {{APPROVER}}.
- If a field is missing, if a file is missing, or if a document is the wrong period, name it, ask once, then stop. Do not proceed on a guess.
- If a figure was not provided, write NOT PROVIDED and name the document it should have come from. Never estimate it.
- Currency: convert to {{BASE_CURRENCY}} and state the rate and its date. Never mix.
- Every agent produces working papers, never a decision. {{APPROVER}} signs.
- Evidence tiers. TIER 1 is a primary document (audited statement, signed agreement, bank file). TIER 2 is a keyed or exported figure (a workbook, an admin export). TIER 3 is MARKETING (a deck, a pitch letter): it makes a question, never a number. Label every load bearing figure REPORTED (with its document and page), CALCULATED (show the arithmetic) or ASSUMED (who set it), plus its tier.
- Before any deliverable leaves the room, run the self check in prompt 12. It blocks.
</shared_rules>

<adapt>
Four blocks change this pack for your own shop, and nothing else needs touching:
- MANDATE changes what "fits" means in every screen.
- EVIDENCE STANDARD raises or lowers what counts as a source. It is the strictest single dial here.
- The desk list in question 1, if your firm names its desks differently.
- The output format of prompt 12, to match the paper your committee already reads.
</adapt>

<trap>People name the artifact they were asked for, not the decision it feeds. "Build me a comparison of these four funds" is usually "I have to defend one of these on Thursday." Route on the second one, or you will hand back the wrong desk.</trap>

<output_format>
Return the run card only: SURFACE (or STOP AND RE ROUTE, with the surface that fits), DESK, STAGE, DATA SOURCE, MODEL (with the one line reason), the five tokens, and the numbers of the three to six agents below that match. Then stop and wait.
</output_format>
Click to copy
Follow the shared rules in prompt 01. This is a practice run on a synthetic sample of four made up jobs. Run the router on each case below, return one run card per case, then compare with the answer key at the end. If any line of your run card does not match, say which line and STOP. Do not use real data until all four match.

CASE A. "I have one manager's DDQ, pitch deck and audited statements, three PDFs, and an IC on Thursday."
CASE B. "Here is a folder export of a data room, about 400 files. I need the full LBO rebuilt from the source documents and every figure checked by a second agent."
CASE C. The human pastes only this line: /Users/me/Desktop/dataroom/
CASE D. "Month end close for three entities. Bank files and ledgers attached for entities 1 and 2. Entity 3's bank file is not here yet."

ANSWER KEY
A. Surface: Claude app, private Project. Desk: F, diligence. Model: Claude Opus 5.5. The stakes are high, but three documents and one manager is not a hard job. Choosing Fable 5.2 here is the classic over pick. The deck is MARKETING, TIER 3: it makes questions, not numbers.
B. Surface: Claude Code, pointed at the folder. Desk: G, deal, with the diligence desk second. Model: Claude Fable 5.2, because it is a full model rebuild from a whole data room with agents checking each other. If the plan has no Fable 5.2, Claude Opus 5.5 with the job split by folder.
C. STOP. A file path in a chat window is a Claude Code job. Say so, name Claude Code as the surface, and do not guess the contents.
D. Surface: Claude app or Claude Code. Desk: B, the close. Model: Claude Opus 5.5. Entities 1 and 2 run now. Entity 3 is OPEN: its bank file is named as missing, and only its reconciliation waits. Holding all three is the wrong answer.
Click to copy
THE CONTROLS DESK, router letter A. Follow the shared rules in prompt 01. Nothing here is released until prompt 12 passes.
Pick by job, not by curiosity. Each link opens that agent's own free page.
<trap>A control that passes on the sample you picked says nothing about the items you skipped. State how the sample was chosen.</trap>
Stop: if the agent you pick is missing the policy, rate card or rule the check runs against. Name it and stop that check only.

001. Agent Governance Kit: Plus AGT install path, three policy YAML templates, OWASP verify runbook.
     https://consultance.ai/library/agent-governance-kit
002. AI Data Residency Check for CFOs: A go/no-go pack for CFOs putting AI on the books.
     https://consultance.ai/library/finance-ai-residency-gate
003. AI Vendor Invoice Audit Kit: Recompute your metered AI invoices against the contract rate card.
     https://consultance.ai/library/ai-vendor-invoice-audit
004. Bank Model Risk Gap Check: Map your model inventory to the new SR 26-2 rule.
     https://consultance.ai/library/model-risk-gap-check
005. Board Variance Bridge Second Read: Run a second read on your board variance bridge.
     https://consultance.ai/library/variance-second-read
006. Finance Team Library Guide: One guide to 51 free packs, mapped to close, cash, spend.
     https://consultance.ai/library/finance-team-library-guide
007. Free AI Build Hardening Kit: Harden what you already built.
     https://consultance.ai/library/builder-hardening-checklist
008. Portfolio Look Through Exposure Audit: See true position overlap across every manager in one exposure map.
     https://consultance.ai/library/allocator-look-through-audit
009. Private Fund Fee Audit Pack: Recompute the management fee against the actual LPA base.
     https://consultance.ai/library/advisor-fee-audit
010. QoE First Read: Rebuild EBITDA from the general ledger, test every add back.
     https://consultance.ai/library/qoe-second-read
011. Second Opinion on Audited Financials: A second review on financials someone already signed.
     https://consultance.ai/library/audit-second-read
012. The Agent Tripwire Pack: Six agent desks that run the sell side pack end.
     https://consultance.ai/library/agent-tripwire-pack
013. The AI Output Verification Pack: Fact-check every citation, quote, and number before you ship it.
     https://consultance.ai/library/ai-output-verification-pack
014. Twenty Finance Desks Control Room: A router that puts any manager or deal job.
     https://consultance.ai/library/twenty-finance-desks
Click to copy
THE CLOSE DESK, router letter B. Follow the shared rules in prompt 01. Nothing here is released until prompt 12 passes.
Pick by job, not by curiosity. Each link opens that agent's own free page.
<trap>A reconciliation that balances can still hide two errors that cancel out. Match in both directions, not just the total.</trap>
Stop: if the agent you pick is missing a bank file or subledger for the period. Name it, mark that account OPEN, and run the other accounts.

015. AI vs Payroll Trade-Off Calculator: A free calculator for CFOs, founders, and ops leads.
     https://consultance.ai/library/ai-capex-reshuffle-calculator
016. CFO AI Close Kit: Run the monthly close in your own Claude.
     https://consultance.ai/library/cfo-ai-close-kit
017. Claude and NetSuite Setup Guide: The setup steps for Oracle's Claude connector plus 12 read only.
     https://consultance.ai/library/netsuite-claude-setup-guide
018. Claude Bookkeeping Stack to Replace QuickBooks: Replace QuickBooks with handle reconciliation, categorisation, AR chasing, and month-end close.
     https://consultance.ai/library/quickbooks-bookkeeping-stack
019. Cut Your Month End Close to One Day: Match the bank to the ledger, book accruals and prepaids.
     https://consultance.ai/library/fast-close-desk
020. Daily Fund NAV Tie-Out: Tie positions, cash, and accruals to the custodian every day.
     https://consultance.ai/library/nav-tieout-engine
021. Finance Back Office Prompt Pack: Draft journal entries, run variance analysis, write the board pack.
     https://consultance.ai/library/customer-zero-finance-ops
022. Free 3-Statement Model Builder: Build a tied-out income statement, balance sheet.
     https://consultance.ai/library/three-statement-model-vault
023. Group consolidation without Hyperion: For group controllers who consolidate several country P&Ls in Excel.
     https://consultance.ai/library/month-end-consolidation
024. In-House Fund NAV Close: Run your monthly fund NAV close in-house.
     https://consultance.ai/library/fund-nav-close-pack
025. Month End Close AI Agent Team: A month end close team in your own Claude that preps.
     https://consultance.ai/library/close-12-agent-team
026. Month End Close Reconciliation Engine: Run the month end close reconciliation itself.
     https://consultance.ai/library/blackline-close-kill
027. Month-End Close and Consolidation Prompts: Run a CPA-grade month-end close and multi-entity consolidation.
     https://consultance.ai/library/cpa-far-bookkeeper-test
028. Month-End Close Automation Pack: Prompts for controllers and CFOs to run the month-end financial close.
     https://consultance.ai/library/month-end-close-kill
029. Oil & Gas Statement Reader: Reads an operator's oil and gas revenue or JIB statement PDF.
     https://consultance.ai/library/oil-gas-non-op-statement-pack
030. Practice AI Bookkeeping On Your Own Books: No CFO trusts AI with the company books on day.
     https://consultance.ai/library/firefly-cfo-trust-ladder
031. Private Asset Reconciliation Engine: Parse notices, match cash to the custodian file, reconcile units.
     https://consultance.ai/library/private-asset-recon-pack
032. Private Ledger Stack: Run reconcile, covenants, redaction.
     https://consultance.ai/library/private-ledger-stack
033. Quarter End Finance Skills for Claude: Run 22 open source finance skills in your own Claude.
     https://consultance.ai/library/quarter-end-finance-skills
034. QuickBooks and Claude Setup Guide: Connect Intuit's QuickBooks connector to Claude, then run for cash, receivables.
     https://consultance.ai/library/quickbooks-claude-setup-guide
035. Ramp Claude Connector Setup Guide: Claude setup steps plus 12 read only prompts for spend baseline.
     https://consultance.ai/library/ramp-claude-setup-guide
036. Replace Concur Expense and AP: Run expense and AP inside Claude instead of Concur.
     https://consultance.ai/library/concur-expense-ap-kill
037. Statement Matching and Reconciliation: Plus an offline extractor that match scanned bank, custodian.
     https://consultance.ai/library/statement-matching-recon
038. Transfer Pricing and Tax Desk: Roll the transfer pricing study forward and flag entities outside.
     https://consultance.ai/library/tax-desk-pack
Click to copy
THE MONITORING DESK, router letter C. Follow the shared rules in prompt 01. Nothing here is released until prompt 12 passes.
Pick by job, not by curiosity. Each link opens that agent's own free page.
<trap>A covenant tested on the borrower's own calculation passes by design. Re derive it from the certificate.</trap>
Stop: if the agent you pick is missing the latest certificate or statement. Mark that position OPEN; never roll last quarter forward.

039. CLO Covenant and Payment Monitor: Read the loan tape and credit agreements.
     https://consultance.ai/library/clo-covenant-monitor
040. Credit Memo AI Workflow Layer: Get credit memos past the two day AI plateau.
     https://consultance.ai/library/past-the-plateau-workflow-layer
041. Free Covenant Monitoring Desk: Recompute leverage and coverage covenants from the borrower's own financials.
     https://consultance.ai/library/covenant-watch-desk
042. Free Fund Manager Monitoring Desk: A dated morning note on every new Form D.
     https://consultance.ai/library/manager-watch-desk
043. Free Portfolio Monitoring Desk: Normalize portfolio company KPIs, flag covenants, and build the LP review.
     https://consultance.ai/library/portco-monitoring-desk
044. Free Portfolio Risk Budget Checklist: A Claude checklist for where your portfolio's risk comes.
     https://consultance.ai/library/portfolio-risk-budget-desk
045. Fund Leverage Read Desk for LPs: Read the debt note before the quarterly letter and see which.
     https://consultance.ai/library/fund-leverage-read-desk
046. Manager Monitoring and Look Through Desk: Read each manager's letter against the memo you funded.
     https://consultance.ai/library/fof-look-through-desk
047. Monthly Portfolio Monitoring Pack: Map each company's monthly file into one template, chase missing numbers.
     https://consultance.ai/library/pe-portfolio-monitoring-pack
048. Morning Portfolio Risk Review in Claude: A before-the-open review of exposure.
     https://consultance.ai/library/analyst-seat-siren
049. Nightly Chart Review for Your Watchlist: And a free engine that review every chart after the close.
     https://consultance.ai/library/chart-desk-night
050. Private Credit Monitoring and Underwriting: Test each certificate against the credit agreement.
     https://consultance.ai/library/private-credit-underwriting
051. Private Credit Surveillance Engine: Monitor covenants, recalculate the borrowing base, read amendment redlines.
     https://consultance.ai/library/moodys-surveillance-kill
052. The Private Credit Agent Stack: The map of ten agent desks that run origination through surveillance.
     https://consultance.ai/library/private-credit-stack
Click to copy
THE RESEARCH DESK, router letter D. Follow the shared rules in prompt 01. Nothing here is released until prompt 12 passes.
Pick by job, not by curiosity. Each link opens that agent's own free page.
<trap>A backtest that beats the market on the data it was tuned on is not an edge. Demand an out of sample test.</trap>
Stop: if the agent you pick is missing a filing or price series for part of the period. Say which part, and do not fill the gap.

053. AI Market Reader Prompt Pack: A free pack of 12 ChatGPT and Claude prompts that read.
     https://consultance.ai/library/market-ai-starter-kit
054. AI stock research and backtest desk: A 14 role AI trading desk modelled on TradingAgents.
     https://consultance.ai/library/quant-desk
055. Blind Earnings Direction Checklist in Claude: A blind checklist over three years of statements that calls next.
     https://consultance.ai/library/earnings-direction-read
056. Claude Market Research Prompt Pack: Claude prompts that run market research on your own question.
     https://consultance.ai/library/gartner-research-synthesis
057. Daily Macro Morning Note: 11 real Claude skills, not copy-paste prompts.
     https://consultance.ai/library/allocator-morning-note
058. Equity Research Desk That Replaces FactSet: A FactSet alternative for equity analysts.
     https://consultance.ai/library/factset-equity-research-kill
059. Explain a Fed Rate Move to Your Board: Turn a rate decision and CPI print into a sourced board.
     https://consultance.ai/library/macro-board-question
060. Free Bloomberg Terminal Alternative: A free, open source terminal for analysts and finance founders.
     https://consultance.ai/library/fincept-terminal
061. Free PME Benchmark Alternative: Rebuild a manager's track record from the dated cash flows.
     https://consultance.ai/library/pme-track-record-rebuild
062. Free Portfolio Quant Research Desk: Backtest an idea on your own portfolio history with free market.
     https://consultance.ai/library/portfolio-quant-desk
063. Hedge Fund Tear Sheet in Excel: Turn your monthly returns into a hedge fund tear sheet.
     https://consultance.ai/library/allocator-tear-sheet
064. IPO Valuation and Comps Builder: Build IPO comps, run the DCF cross-check.
     https://consultance.ai/library/ipo-comp-models
065. Manager Backtest Stress Test Checklist: A checklist that splits built years from unseen ones.
     https://consultance.ai/library/manager-backtest-stress-test
066. Manager Benchmark Mismatch Check: Rebuild a manager's real exposure against the stated benchmark.
     https://consultance.ai/library/allocator-benchmark-desk
067. Personal AI Stock Research Desk: Argue one stock long and short in your own Claude.
     https://consultance.ai/library/vibe-trading-personal-desk
068. Private Equity Deal Sourcing in Claude: Run deal sourcing and target screening in Claude with 12 tested.
     https://consultance.ai/library/pe-analyst-confession
069. Private Equity Deal Sourcing Playbook: Turn one mandate into a ranked.
     https://consultance.ai/library/deal-origination-engine
070. Private Equity Diligence and Deal Toolkit: Turn Claude into a private equity diligence and deal toolkit.
     https://consultance.ai/library/kkr-style-guide-to-claude
071. Private Equity Diligence Prompt Pack: Run private equity deal diligence.
     https://consultance.ai/library/pe-lbo-diligence-kill
072. SEC Filings Reader in Claude: Answer from 10-Ks, 8-Ks, Form 4s and 13Fs.
     https://consultance.ai/library/sec-filings-reader
073. Sector Overview Engine: Build a full sector overview from public filings, market size, players.
     https://consultance.ai/library/sector-overview-engine
074. Shared Desk Terminal for Buy-Side Teams: Plus 4 bonus that turn one Claude into a desk terminal.
     https://consultance.ai/library/desk-terminal
075. Stock Research Structured Review With Fact Check: A structured review of one listed stock in Claude.
     https://consultance.ai/library/research-memo-fact-check
076. The Private Equity Agent Stack: The map of ten agent desks that run the deal.
     https://consultance.ai/library/pe-stack
077. Trading Comps Builder in Claude: Screen the peer set, calendarize the fiscal years.
     https://consultance.ai/library/trading-comps-vault
Click to copy
THE FP&A DESK, router letter E. Follow the shared rules in prompt 01. Nothing here is released until prompt 12 passes.
Pick by job, not by curiosity. Each link opens that agent's own free page.
<trap>Annualising a seasonal first half gives a clean and wrong number.</trap>
Stop: if the agent you pick is missing actuals and plan that tie at the start. Stop the forecast until they tie; other reports still run.

078. Board Strategy Deck Prompt Pack: Turn your own data into a board ready strategy deck.
     https://consultance.ai/library/mckinsey-strategy-dashboard
079. CFO 13 Week Cash Forecast Kit: Put a second pair of eyes on the 13 week cash.
     https://consultance.ai/library/cfo-cash-forecast-kit
080. Claude Dashboards for CFOs: Turn your Excel or accounting numbers into a chart.
     https://consultance.ai/library/power-bi-claude-dashboards
081. Daily Cash Position From Bank Files: Turn MT940, BAI2 and CAMT.053 files into one cash position.
     https://consultance.ai/library/bank-file-cash-position
082. Financial Statement Preparation Pack: Map the trial balance, draft the notes.
     https://consultance.ai/library/finance-reporting-pack
083. FP&A Rolling Forecast Engine: Rebuild the rolling forecast off your actuals the day the close.
     https://consultance.ai/library/fpna-forecast-engine
084. Live Reports That Replace Tableau: Turn raw finance numbers into a live.
     https://consultance.ai/library/live-reporting-pack
085. LP Capital Call Forecast Desk: Forecast next quarter's capital calls off your own fund history.
     https://consultance.ai/library/lp-capital-call-pacing-desk
086. Quarterly Client Report Builder: Build the quarterly multi asset client report.
     https://consultance.ai/library/wealth-client-report-pack
087. The FP&A Agent Stack: The map of ten agent desks that run the planning cycle.
     https://consultance.ai/library/fpna-stack
088. Treasury Agent Stack: Covering cash position, payment review, forecast refresh, covenant watch.
     https://consultance.ai/library/treasury-agent-stack
Click to copy
THE DILIGENCE DESK, router letter F. Follow the shared rules in prompt 01. Nothing here is released until prompt 12 passes.
Pick by job, not by curiosity. Each link opens that agent's own free page.
<trap>The seller's add back schedule is a claim, not a figure. Test each one against a document.</trap>
Stop: if the agent you pick is missing a data room document. List it as OPEN and keep reading the rest of the room.

089. AI Deal Diligence Prompt Pack: Read a data room and draft the IC memo in Claude.
     https://consultance.ai/library/deal-room-autopilot
090. AI Portfolio Manager Prompt Pack: Write the investment policy statement, set the asset allocation.
     https://consultance.ai/library/cfa-level3-ai-test
091. Allocator Excel Desk: Demo the desk on a public sample you can show.
     https://consultance.ai/library/allocator-excel-desk
092. Allocator Red Flag Desk: Cross the DDQ against the audited financials.
     https://consultance.ai/library/allocator-red-flag-desk
093. Claude Diligence Desk: From mandate screen to IC memo, with EBITDA.
     https://consultance.ai/library/claude-diligence-desk
094. Commercial Due Diligence Desk: Index the data room, size the market.
     https://consultance.ai/library/bain-cdd-diligence-kill
095. Data Room Cost Split: Map the room by deal stage.
     https://consultance.ai/library/datasite-papermark-split
096. Deal Red Flag Structured Review: A structured review that sorts signed from draft.
     https://consultance.ai/library/dd-red-flag-deal-action
097. Free Quality of Earnings Toolkit: Run quality of earnings due diligence yourself, EBITDA addbacks included.
     https://consultance.ai/library/quality-of-earnings-desk
098. Insurance Agency Diligence Review: Break down the book, split growth into lost, rate.
     https://consultance.ai/library/insurance-agency-diligence
099. LP Due Diligence Questionnaire Desk: Match a new LP questionnaire to your past answers.
     https://consultance.ai/library/lp-ddq-desk
100. Manager Due Diligence Pack: Tie returns to the audit, read the letters.
     https://consultance.ai/library/allocator-manager-dd
101. Manager Search and RFP Review Prompts: Claude prompts that score RFP answers against your policy and check.
     https://consultance.ai/library/manager-search-desk
102. Open Source PE Deal Workspace on Claude: Run an open source deal workspace on your own machine.
     https://consultance.ai/library/goodbye-dealcloud-pe-workspace
103. Outside-In Diligence Review: Test every CIM claim against pricing history, hiring, reviews.
     https://consultance.ai/library/outside-in-diligence
104. PE Diligence Pack for Deal Teams: Underwrite a mid-market buyout end to end, the data room read.
     https://consultance.ai/library/pe-diligence-pack
105. Side Letter Review Desk for Allocators: Read the LPAs and side letters you've already signed.
     https://consultance.ai/library/allocator-side-letter-desk
106. Spot Accounting Red Flags: Turn Claude into a forensic accountant for any financial statement.
     https://consultance.ai/library/forensic-skeptic-pack
107. Startup Code Review for Investors: Have your own Claude read the founder's code.
     https://consultance.ai/library/startup-code-diligence
108. Startup Diligence Pack for VCs: Run the diligence a VC partner uses to kill a deal.
     https://consultance.ai/library/vc-startup-diligence-pack
109. Submission to Quote Underwriting Pack: Read a submission, pull the loss runs, build the technical price.
     https://consultance.ai/library/insurance-underwriting-pack
110. The Blind Underwrite Kit: The map of six agent desks that run a second underwrite.
     https://consultance.ai/library/blind-underwrite-kit
111. The Deal Model Red Flag Checklist: A one page checklist of 15 places a deal model hides.
     https://consultance.ai/library/diligence-red-flag-checklist
Click to copy
THE DEAL DESK, router letter G. Follow the shared rules in prompt 01. Nothing here is released until prompt 12 passes.
Pick by job, not by curiosity. Each link opens that agent's own free page.
<trap>A model that ties to itself can still not tie to the source. Check the inputs before the outputs.</trap>
Stop: if the agent you pick is missing the term sheet or the latest financials. Name them and do not price the deal.

112. AI Agent P&L Board Template: Price every AI finance agent against the salary it replaces.
     https://consultance.ai/library/agent-pnl-board-pack
113. Anthropic vs OpenAI Valuation Model: Plus a live Anthropic vs OpenAI dashboard that builds a sourced.
     https://consultance.ai/library/anthropic-vs-openai-comp-vault
114. Build a CIM in Claude: Anthropic's free finance plugin drafts the CIM from your data room.
     https://consultance.ai/library/cim-builder
115. Cap Table Rebuild and Waterfall Desk: Rebuild it from the signed documents, name every share without paper.
     https://consultance.ai/library/cap-table-desk
116. Continuation Fund Election Desk for LPs: Rebuild the portco's value from its own last three quarterly reports.
     https://consultance.ai/library/continuation-vehicle-election-desk
117. Free DCF Model Builder: Build a full discounted cash flow model on your own numbers.
     https://consultance.ai/library/dcf-model-builder
118. Free Jira Alternative for Deal Teams: A self hosted deal tracker your Claude can write.
     https://consultance.ai/library/agent-writable-deal-tracker
119. Free Sell Side Deal Desk: Draft the teaser, the CIM, and the buyer list in Claude.
     https://consultance.ai/library/sell-side-deal-desk
120. Free Sell-Side Teaser Kit: Turn raw financials into the anonymized 2-page M&A teaser a banker.
     https://consultance.ai/library/sell-side-teaser-kit
121. IC Memo Bull Bear and Kill Line: Write the investment committee memo, the bull case, the bear case.
     https://consultance.ai/library/ic-memo-desk
122. LBO Model and Returns Calculator: Build an LBO model in an afternoon with 8 prompts.
     https://consultance.ai/library/lbo-returns-model
123. Private Equity Co Invest Screening Desk: An independent base case, add-back test, downside.
     https://consultance.ai/library/coinvest-screening-desk
124. Purchase Price True Up Desk: Rebuild closing working capital from the ledger.
     https://consultance.ai/library/purchase-price-bridge-desk
125. Valuation First Pass Pack: Check the DCF against the comps, rebuild comps, precedents and sensitivities.
     https://consultance.ai/library/dcf-attack-kit
Click to copy
THE INVESTOR DESK, router letter H. Follow the shared rules in prompt 01. Nothing here is released until prompt 12 passes.
Pick by job, not by curiosity. Each link opens that agent's own free page.
<trap>A capital account that differs from the GP statement is often a timing difference, not an error. Check the period before you flag it.</trap>
Stop: if the agent you pick is missing the LPA or a side letter. Say so, and do not infer fee terms from the PPM.

126. Family Office CRM Alternative In Excel: Track deal flow, managers and call notes in the sheet you.
     https://consultance.ai/library/family-office-tracker-desk
127. Family Office Quarterly Report: Roll every custodian statement into one book, reconcile entities.
     https://consultance.ai/library/family-office-investment-pack
128. Free Big 4 Advisory Agent Kit: Three Claude agents that monitor tax and regulation exposure.
     https://consultance.ai/library/kpmg-advisor-agent-swap-kit
129. Free Family Office Meeting Pack: Build the one page family meeting brief overnight.
     https://consultance.ai/library/family-meeting-pack
130. Free Hedge Fund Risk Desk: A morning risk brief from your own position export.
     https://consultance.ai/library/morning-risk-pod
131. Free TAMP Alternative Desk: Run household drift checks, tax loss harvesting scans.
     https://consultance.ai/library/ria-advisor-desk
132. GP Fundraise Prep Pack: An LP target list ranked by room this vintage.
     https://consultance.ai/library/gp-fundraise-prep
133. LP Capital Account Tie Out Desk: Tie every GP capital account statement to your own cash.
     https://consultance.ai/library/lp-capital-account-tieback-desk
134. Portfolio AI Maturity Scan: Point your own Claude at a company and score where.
     https://consultance.ai/library/portfolio-ai-maturity-scan
135. Quarterly LP Reporting Pack: Tie every LP capital account before the quarterly letter ships.
     https://consultance.ai/library/lp-quarterly-pack
136. RIA Practice Strategy Skills for Claude: For practice diagnosis, client segmentation, fee tier design and growth roadmaps.
     https://consultance.ai/library/advisor-strategy-desk
137. The Family Office Agent Stack: The map of ten agent desks that run the mandate.
     https://consultance.ai/library/family-office-stack
Click to copy
THE STACK BUILDERS, router letter I. Follow the shared rules in prompt 01. Nothing here is released until prompt 12 passes.
Pick by job, not by curiosity. Each link opens that agent's own free page.
<trap>More agents is not more coverage. Install the one that covers a job nobody does today.</trap>
Stop: if the agent you pick is missing the system the agent will read from. If the human cannot name it, ask before building anything.

138. Accretion Dilution Merger Model: Run accretion or dilution, the funding mix, goodwill, synergies.
     https://consultance.ai/library/merger-model-kit
139. AI Discovery Playbook for Leaders: Run the enterprise AI discovery a consultancy charges $1M.
     https://consultance.ai/library/ai-discovery-playbook
140. AI Investment Committee for Stocks: Install the open ai-berkshire repo and turn your own Claude Code.
     https://consultance.ai/library/ai-investment-committee
141. Big 4 Consulting Prompts for Claude: Finance and ops teams run for Big 4 consulting workflows like.
     https://consultance.ai/library/kpmg-big4-prompt-vault
142. Big 4 Deliverables Prompt Pack: Run six Big 4 grade deliverables yourself with 12 prompts.
     https://consultance.ai/library/big4-judgment-gate
143. Build Your Own AI Agent Desk: Run ai-hedge-fund, the open source AI hedge fund, on Claude.
     https://consultance.ai/library/build-your-agent-desk
144. Claude Code Trading Agents for Robinhood: Run Claude Code trading agents on your Mac through Robinhood's MCP.
     https://consultance.ai/library/opentrade-agent-trading-pod
145. Claude finance prompts for Excel: Tie a three statement model in Excel, forecast cash.
     https://consultance.ai/library/claude-m365-finance-pack
146. Complete Guide to AI in Finance: Ten verified papers and tools, each with what it shows.
     https://consultance.ai/library/finance-ai-complete-guide
147. Finance Skills for Claude Code: Install 7 open source finance skills in Claude Code for valuation.
     https://consultance.ai/library/goodbye-factset-finance-skills
148. Finance Skills OS for Claude: The diligence desk an EY or PE team spins up.
     https://consultance.ai/library/finance-skills-os
149. Firm AI Self-Diagnosis Checklist: A self-review checklist that prices your recurring work.
     https://consultance.ai/library/firm-ai-self-diagnosis
150. Free 42 Agent Finance Org Pack: Install 42 ready made Claude Code agents, a CFO.
     https://consultance.ai/library/finance-42-agent-org
151. Free Karpathy LLM Council for Claude: A structured decision review in your own Claude.
     https://consultance.ai/library/claude-decision-council
152. Free Open Source Finance AI Tools: Ten open source finance AI tools mapped to five research jobs.
     https://consultance.ai/library/finance-ai-repo-playbook
153. Investment Banking AI Agent Team: 10 ready made Claude Code agents for origination.
     https://consultance.ai/library/ib-10-agent-deal-team
154. Open Source AI Finance Research Agent: A free open source AI agent that pulls live quotes, filings.
     https://consultance.ai/library/opencandle-install-guide
155. PE Ops Intel Screen From Exports: Inventory exports, score each function against the plan you underwrote.
     https://consultance.ai/library/ops-intel-screen
156. PE Portfolio Automation Agents: Find which portfolio company workflows an AI agent can take over.
     https://consultance.ai/library/pe-portfolio-automation-agents
157. Post Earnings Model Update Kit: Roll the quarter, reset the run rate.
     https://consultance.ai/library/earnings-update-kit
158. Private Equity AI Agent Team: 10 Claude Code agents for sourcing.
     https://consultance.ai/library/pe-10-agent-fund-team
159. RIA Back Office AI Agent Team: 10 ready made Claude Code agents that work as a screening.
     https://consultance.ai/library/ria-10-agent-advisor-desk
160. RIA Operations Agent Team: Check new account paperwork, track transfers, tie out the custodian.
     https://consultance.ai/library/ria-operations-agent-org
161. Run Your Finance Team in Slack: Run your month-end close, AP.
     https://consultance.ai/library/claude-tag-finance-team
162. Seven Finance Agents That Run Overnight: Stand up seven finance analysts as scheduled agents.
     https://consultance.ai/library/always-on-finance-analyst
163. Strategy Consulting Skills for Claude: Run a full strategy engagement inside your own Claude.
     https://consultance.ai/library/pe-strategy-skill-pack
164. Technology Vendor Quote Teardown Kit: Split a technology vendor quote into the part AI turned.
     https://consultance.ai/library/vendor-quote-teardown
165. The CFO Agent Stack: The map of six agent desks that run the close.
     https://consultance.ai/library/cfo-agent-stack
166. The CFO Claude Stack: The map of seven finance desks, deal, diligence, close, tax, underwriting.
     https://consultance.ai/library/cfo-claude-stack
167. The FRM Part II Risk Judgment Pack: Run the applied risk judgment calls on your own numbers.
     https://consultance.ai/library/frm-risk-judgment-pack
168. The Investment Bank Agent Stack: The map of ten agent desks that run the sell side.
     https://consultance.ai/library/investment-bank-stack
169. Which AI model for each finance job: A free routing map for which AI model wins.
     https://consultance.ai/library/finance-ai-model-map
Click to copy
You are an independent reviewing manager. You did not do the work above and you are not trying to defend it.

<task>
Re derive the load bearing figures from the source documents, independently, before reading anything the run produced. Then compare.
Check: does every figure tie to a document supplied in prompt 01's data source, at the page cited. Is every unsourced figure labelled ASSUMPTION. Is every NOT PROVIDED still marked NOT PROVIDED rather than quietly filled. Did any agent above proceed past a missing document instead of stopping.
</task>

<output_format>
PASS or BLOCK on the first line, nothing before it.
Then one row per check: check, result, evidence, and for a fail, the specific figure and document.
</output_format>

<constraints>
Obey the shared_rules block in prompt 01. This agent appends nothing to the deliverable and softens nothing.
On BLOCK, do not produce the deliverable. Return the fail list only.
BLOCK is cleared one way only: {{APPROVER}}, a named human, resolves the specific discrepancy or accepts it in writing on the record. It is not cleared by re running this prompt, and it is not cleared by an argument that the discrepancy does not matter.
Constructing a case for why a fail is immaterial is itself the failure mode. Do not do it.
</constraints>

<trap>A reviewer that reads the run's own output first will agree with it. Re derive from the source documents before opening anything the run produced.</trap>

<review_gate>This is the gate. Nothing sits above it except {{APPROVER}}.</review_gate>

From a working build to an operating system.

For CFOs, controllers, allocators and PE and IB teams: 169 named finance AI agents by desk, each linking to its free pack

Path A · free

Run the build independently

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
Path B · done with you

Connect it to your operation

We would wire the shelf into your tenant: the router pointed at your own internal desks alongside the published ones, governed connectors to your ERP, administrator and data room so the evidence assembles itself, your evidence standard encoded as tests, and the self check set as a standing gate before anything reaches a board or an investment committee. Done with you, then handed over so you own it.

Book a build call
data safety

Before you use live numbers

  • • Run last quarter's numbers first. Live data is not a test bed.
  • • Nothing here uploads to us. It runs in your own Claude account, on your own machine.
  • • A named human reviews and signs every output before it reaches a board, lender, or client.
  • • Mask account numbers and names to the minimum the task needs.
the fine print

Straight answers on ownership

Prompt set authored by consultance.ai. It is an index over packs already published free on this site and replaces no named product. Your documents stay in your own Claude tenant; we never see them. The agents produce working papers for a named human to sign, not a valuation, an audit opinion, a tax filing, or investment, audit or legal advice.

Bring this workflow into your operation.

Book a build callBack to the library

Want this wired into your stack instead of running it yourself? That is our AI deal desk and finance automation service.

Newsletter

New builds, as they ship

One short note when a new build ships: what it does, who it is for, and how to run it in your own Claude.

Free. Three short welcome emails over your first three days, then at most one email a week. Unsubscribe in one click.

More like this

Related builds

All builds
Finance and data
Finance and data

115 Finance AI Agents Directory

For CFOs, controllers, allocators and PE and IB teams: 115 named finance AI agents mapped to their desk, each linking to its own free pack. A 4 question router narrows it to your 3 to 6.

Setup guide
Twenty Finance Desks Control Room coverFinance and data
Finance and data

Twenty Finance Desks Control Room

For allocators and family offices: a 12 prompt router that puts any manager or deal job on the right desk and blocks the memo when figures do not tie. Opens all 221 named agents across the 20 published packs, free.

Setup guide
Finance and data
Finance and data

Which AI model for each finance job

A free routing map for which AI model wins each finance job: Claude for diligence, GPT for sourcing, Gemini for deal marketing. You get a starter prompt for every lane.

Setup guide
Get the free kitBook a call

Share with your workflow owner.

Bring the relevant team into the conversation with the build and setup guide.

Forward by email
in one line

What is 169 Finance AI Agents Directory?

169 Finance AI Agents Directory is a finance and data build in the consultance.ai AI Build Library. For CFOs, controllers, allocators and PE and IB teams: 169 named finance AI agents by desk, each linking to its free pack. The router picks your 3 to 6 and sends hard jobs to Fable 5.2. It fits CFOs and fund CFOs, controllers, allocators and family office principals, PE and investment banking teams who want the whole shelf in one file and a router that stops them running the wrong three agents. Setup difficulty is Medium, with 4 plain-English steps.

What does 169 Finance AI Agents Directory do?

For CFOs, controllers, allocators and PE and IB teams: 169 named finance AI agents by desk, each linking to its free pack. The router picks your 3 to 6 and sends hard jobs to Fable 5.2.

Who is 169 Finance AI Agents Directory for?

It fits CFOs and fund CFOs, controllers, allocators and family office principals, PE and investment banking teams who want the whole shelf in one file and a router that stops them running the wrong three agents.

How hard is 169 Finance AI Agents Directory to set up?

Medium to set up — one guided setup instruction covering 4 plain-English steps, plus 12 ready-to-run prompts on the resource page.

How would consultance.ai build this out?

We would wire the shelf into your tenant: the router pointed at your own internal desks alongside the published ones, governed connectors to your ERP, administrator and data room so the evidence assembles itself, your evidence standard encoded as tests, and the self check set as a standing gate before anything reaches a board or an investment committee. Done with you, then handed over so you own it.

What are the licensing terms?

Prompt set authored by consultance.ai. It is an index over packs already published free on this site and replaces no named product. Your documents stay in your own Claude tenant; we never see them. The agents produce working papers for a named human to sign, not a valuation, an audit opinion, a tax filing, or investment, audit or legal advice.

Connect this build to your workflow.

Discuss how 169 Finance AI Agents Directory fits your systems, controls, and team. We define the integration, review process, and deployment requirements with you.

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.

Book an audit