An AI Skill for the A.CRE Commercial Mortgage Loan Analysis Model
I’ve been working on a project to make our library of Excel models AI-ready. The idea is straightforward: pair every A.CRE Excel model with an AI Skill, a packaged set of instructions and reference files that teaches an AI assistant how to operate that specific model on your behalf. The Commercial Mortgage Loan Analysis Model is the latest in that effort, and this post introduces the AI Skill we built to accompany it.
Think of this as a sister post to the Commercial Mortgage Loan Analysis Model post, which walks through the model itself, sections, inputs, outputs, and mechanics. If you haven’t seen that one yet, start there. This post focuses specifically on the AI Skill: what it does, how it works, and how to use it.
- While we refer to these as Claude Skills (or AI Skills, Agent Skills), the instructions inside the Skill are largely platform-neutral. You can use the Skill with Claude in other platforms such as ChatGPT, Gemini, or any other capable AI assistant. You can also use these in Claude in Excel, ChatGPT in Excel, etc.
What is an AI Skill?
If you’re new to the concept, an AI Skill is a packaged set of instructions and reference files that an AI assistant loads alongside your file. It teaches the assistant things it wouldn’t otherwise know — in this case, every input cell, every output, the four user roles the Commercial Mortgage Loan Analysis Model serves, the mechanics behind the four loan-sizing tests, and the most common ways analysts misconfigure the model and produce wrong loan amounts.
The result is an AI assistant that can actually operate the model on your behalf, rather than one that talks about commercial mortgage underwriting in the abstract.
For a primer with a short video tutorial, see our practical guide to Claude Skills.
What the AI Skill Does for You
Commercial mortgage analysis looks deceptively similar across the four user roles — borrower, lender, broker, student — but the investment question each is answering is completely different. A borrower wants the most debt they can get and what it actually costs, inclusive of the loan fee, not just the stated rate. A lender wants to know if the loan request clears the credit box and where the binding constraint sits. A broker wants to run multiple structures side by side and match each to the right lender pool. A student wants the mechanics explained from first principles. The model handles all four, but the right inputs, the right outputs, and the right framing differ materially in each case.
So the Skill handles a few jobs for you that you’d otherwise be doing manually.
Role Triage
Before the Skill touches a single input, it asks which side of the loan you’re on:
- Borrower / sponsor — sizing a new acquisition or refinance loan. The Skill surfaces the Maximum Loan Amount, which test is binding, and the Lender Yield (APR), framed as the most debt you can get and what it actually costs you.
- Lender / debt fund — underwriting a loan request against a stated credit box. The Skill surfaces pass/fail flags for each test, the binding constraint at the borrower’s requested amount, and cushion above the floors — because the lender’s job is understanding downside buffer, not maximizing loan size.
- Mortgage broker — advising on debt placement and capital structure. The Skill runs multiple structures side by side, explains which test binds under each, and surfaces the Lender Yield and Average Life that determine which lender pool each structure fits.
- Student / Accelerator member — learning how commercial mortgage sizing works. The Skill walks through each of the four sizing tests individually, builds from the Pro Forma NOI up through the max loan logic, and explains why APR and note rate are two different numbers.
Populating Inputs Conversationally
You can paste in a term sheet, upload an offering memorandum, share a T-12 operating statement, or describe the property and proposed loan terms in plain language — the Skill pulls the relevant details (property type, NOI, cap rate, loan structure, rate components, test floors) and stages them for the model. One step it always confirms before writing: the Pro Forma Underwriting column (column Q), which drives all downstream sizing, versus the T-12 (column M). These are different numbers and the distinction matters. The Skill flags this every time.
It also confirms the four sizing test toggles before reporting the Maximum Loan Amount. These default to “Yes” for all four tests, but different lenders apply different combinations — a balance-sheet bank for stabilized multifamily typically doesn’t enforce a debt yield floor, while a life insurance company lender often leads with it. Setting the wrong toggles means the max loan is sized against tests the lender doesn’t actually apply.
Catching Common Mistakes
Six errors show up repeatedly on this model. The Pro Forma NOI populated from the T-12 column rather than the underwriting column, silently sizing the loan on the wrong base. A stale benchmark rate on the Loan Terms tab, which ships with a placeholder and moves daily. The sizing test toggles left at their defaults without confirming which tests the lender actually applies. The Proposed Loan Amount left at the formula default when the user actually wants to size to a specific LTV or dollar target. The Lender Yield confused with the stated note rate — these are different numbers, and the difference is the loan fee amortized over the term. And under floating-rate mode, the SOFR forward curve on the Amortization tab not populated or refreshed, which makes the amortization schedule and DSCR outputs unreliable.
Framing Outputs in Your Role
The headline output for a borrower is the Maximum Loan Amount paired with which test binds — because that tells you exactly which lever to pull if you need more proceeds. The headline output for a lender is the pass/fail stack with cushion above each floor — because that is the downside buffer narrative. The headline for a broker is a comparison table across structures, matched to lender pool. The Skill frames these differently because the same number means something different depending on who’s reading it.
One output framing note that applies across all roles: Lender Yield (APR) is the true cost of capital to the borrower, and the true yield to the lender. It includes the loan fee amortized over the loan term. The stated note rate does not. The Skill surfaces this distinction the first time either number comes up.
Operating Contexts (Chat / Cowork and Claude in Excel)
The Skill works in two environments. You can upload the Excel file to a Claude conversation and have Claude operate the model via code execution — the Skill bundle includes a clean copy of the model so no upload is needed to get started. Or, if you’re using Claude in Excel, operate the model live with Claude reading and writing to the workbook directly. The Skill handles both, including the refinance risk sensitivity table and the prepayment fee calculation, with mechanics adjusted appropriately under the hood. And as noted earlier, the Skill is also portable to other AI assistants, though the integration may be lighter.
A Note on the Underlying Model
The Commercial Mortgage Loan Analysis Model is a property-type-agnostic permanent-loan tool for underwriting and sizing commercial mortgage debt. It runs four loan-sizing tests — DSCR Amortizing, DSCR Interest-Only, Debt Yield, and LTV — individually toggleable, against a direct-cap property pro forma to compute the Maximum Loan Amount. It also outputs Lender Yield (APR), Weighted Average Life, a refinance balloon-risk sensitivity table across NOI and cap rate scenarios, and a prepayment fee analysis using either Yield Maintenance or a percentage-of-balance method. Fixed and floating rate debt are both supported. The model covers six tabs: Loan Summary, Pro Forma, Loan Terms, Refinance Analysis, Prepayment Analysis, and Amortization. See the model post for the full walkthrough.
Note: This AI Skill is built for v2.7 of the model. If you’re on an older version, the Skill will flag the mismatch — notably v2.4 introduced the Fixed vs Floating toggle and the variable rate module, and v2.5 changed the DSCR Amort Test and Maximum Loan Amount formulas. You’ll want the current version of both the model and the Skill for the cleanest experience.
Commercial Mortgage Loan Analysis Model + AI Skill Video Walkthrough: Using the AI Skill
The video below walks through the full AI-assisted workflow: uploading the model, selecting your role, populating the Pro Forma and Loan Terms inputs from a deal summary or term sheet, confirming the sizing test toggles, and interpreting the Maximum Loan Amount, Lender Yield, and binding constraint in the language of the investment decision.
Before You Use This AI Skill with the Commercial Mortgage Loan Analysis Model
A couple of notes worth surfacing before you download.
Who this Skill is for. This Skill is built for real estate professionals with a strong grasp of financial modeling, and ideally prior exposure to commercial mortgage underwriting, loan sizing and structuring, and real estate capital markets. It’s best suited to graduates of our A.CRE Accelerator real estate financial modeling program, or analysts comfortable building debt models from scratch. AI assistants make mistakes; the Skill assumes an analyst on the other side who can catch them. Treat its output the way you’d treat work from a sharp junior analyst, useful, fast, and always verified before it goes into a credit memo.
License. The Skill is distributed under the A.CRE software license, with full terms in the LICENSE.txt file included in the bundle. The short version: use it for personal, organizational, and client-facing analysis; don’t resell or redistribute it. Use by an AI assistant operating on your behalf is expressly permitted, that’s the whole point.
Download the Commercial Mortgage Loan Analysis Model + AI Skill
To make this model accessible to everyone, it is offered on a “Pay What You’re Able” basis with no minimum (enter $0 if you’d like) or maximum (your support helps keep the content coming – typical real estate Excel models sell for $100 – $300+ per license). Just enter a price together with an email address to send the download link to, and then click ‘Continue’. If you have any questions about our “Pay What You’re Able” program or why we offer our models on this basis, please reach out to either Mike or Spencer.
Your download includes three files: the Excel model, the AI Skill (.skill file), and a short README explaining how to use them together.
We regularly update both the model and the AI Skill (see version notes below). Paid contributors receive a new download link via email each time either is updated.
Frequently Asked Questions about an AI Skill for the A.CRE Commercial Mortgage Loan Analysis Model
Version Notes — AI Skill
Version 2.7
- Initial release of the AI Skill for the A.CRE Commercial Mortgage Loan Analysis Model
- Paired with v2.7 of the Excel model
- Supports both Chat / Cowork (Skill bundle includes a clean copy of the model — no upload needed) and Claude in Excel (operate the live workbook directly)
- Includes 4-role triage (borrower/sponsor, lender/debt fund, mortgage broker, student/Accelerator member), conversational input population from term sheets and T-12s, and mistake-catching across the Pro Forma underwriting column, sizing test toggles, stale benchmark rate, Lender Yield vs note rate distinction, and floating-rate SOFR curve state
- Portable to other capable AI assistants (ChatGPT, Gemini, etc.) via the SKILL.md file
