--- name: al-542-screen-every-commercial-property-offering-memo-against-yo description: "In a May 2026 tutorial for AI for Real Estate HQ, Zakee, who runs the proptech platform Buildable, shows how to turn a first-pass commercial underwrite into a reusable Claude Skill. You put three of your own files in a folder: a written buy box (strategy, markets, deal size, return targets, standard assumptions and hard fails), your firm's real underwriting spreadsheet and your rent and sales comps. You ask Claude to build the skill and to ask you design questions first. His answers: never invent missing figures (write 'Not stated. Assumption: X. Rationale: Y.' and make the user confirm), cite the memo page for every number, stop at once on hard fails, and keep sensitivity to a simple table that moves the exit cap rate up and down. After that, dropping in an offering memo and typing 'Underwrite this deal' gives a buy box screen, the in-place figures with page citations, returns at the asking price, the offer price that meets each return target and a short investment memo as a Word…" --- # Screen every commercial property offering memo against your buy box in 20 minutes: build an AI skill once from your criteria, your own model and your comps, then drop in a memo to get pass or proceed, returns at the asking price and the offer price that hits your target In a May 2026 tutorial for AI for Real Estate HQ, Zakee, who runs the proptech platform Buildable, shows how to turn a first-pass commercial underwrite into a reusable Claude Skill. You put three of your own files in a folder: a written buy box (strategy, markets, deal size, return targets, standard assumptions and hard fails), your firm's real underwriting spreadsheet and your rent and sales comps. You ask Claude to build the skill and to ask you design questions first. His answers: never invent missing figures (write 'Not stated. Assumption: X. Rationale: Y.' and make the user confirm), cite the memo page for every number, stop at once on hard fails, and keep sensitivity to a simple table that moves the exit cap rate up and down. After that, dropping in an offering memo and typing 'Underwrite this deal' gives a buy box screen, the in-place figures with page citations, returns at the asking price, the offer price that meets each return target and a short investment memo as a Word document. He is clear about the limits: it replaces the first hour of screening, not the comps, the broker call, the site visit or the full model you build once a deal passes, and it uses a lot of AI usage. ## Inputs to collect first Ask the user for anything below that they haven't already given you. Don't guess missing facts. - Criteria, markets, deal size, return targets, standard assumptions… - Your own material (text, notes or data) ## Instructions Do the task below with the user's inputs in place of each [bracket]. I want to build a reusable skill that screens property offering memos (OMs). I'm attaching three files: my buy box [criteria, markets, deal size, return targets, standard assumptions, hard fails], my underwriting spreadsheet and my comps file. Before building anything, ask me 5-7 design questions about how it should behave. Build in these rules whatever I answer: - Never invent a figure. If the OM doesn't state it, write "Not stated. Assumption: [X]. Rationale: [Y]." and ask me to confirm before continuing. - Cite the OM page for every number you pull. - If the deal fails a hard fail in my buy box, stop and say which one. - If the property's submarket isn't in my comps file, flag it. For each OM, produce: pass or proceed with reasons; the in-place figures (price, NOI, cap rate, units, year built, occupancy); my returns at the asking price using my assumptions; the highest offer price that still meets each of my return targets; a small table moving the exit cap rate [0.5]% up and down; and a one-page investment summary. End with the open questions I should put to the broker. This is a first-pass screen only; I'll build the full model myself for deals that pass. ## Safety checks (share these with the user) - Check every figure, price and total against your own records before anything is sent, posted or booked. A person approves refunds, discounts, payments and price changes. - Check names, dates, numbers and any claim against your own source material before you use the answer. AI can misread or invent details. - Say clearly which parts need a person to review or approve before anything is sent, paid, posted or filed. ## About this skill - Works with: Claude, Claude Skills, Excel - Based on this real-world example: https://aiforrealestatehq.substack.com/p/build-a-claude-skill-that-underwrites - Free prompt from AgentPlaybook Real Estate: https://agentic-library.netlify.app/prompt/al-542-screen-every-commercial-property-offering-memo-against-your/ - To install in an agent that supports Agent Skills, save this file as SKILL.md in a folder with the same name as `name` above (for example in your skills directory). Install steps for Claude.ai, Claude Code, Cursor and Codex: https://agentic-library.netlify.app/skills/