MCP Use Cases — Deal Screening, Offers, Comps & STR Analysis in Your AI

Docs navigation

Screen deals without leaving the chat

Paste a listing's numbers mid-conversation; your AI underwrites it on the real engine instead of estimating. Score, cash flow, stress tests, and a shareable report in one turn.

claude — deal screening

> Zillow listing: $285K duplex, taxes $3,200, should rent ~$3,100/mo.
  I'd put 25% down at 6.75%. Worth a showing?

Called analyze_deal(strategy: "ltr", …)

96/100 Strong — $637/mo cash flow, survives all 7 stress tests.
Yes, book the showing. Full report: reizer.io/report.html?id=…

Prompts that work verbatim:

Underwrite this rental: $285K purchase, $3,100/mo market rent, 25% down at 6.75%. Worth it?
Same deal, but what if rates hit 8% and vacancy doubles? (stress tests are built into every analysis)
Send me the report link so I can forward it to my lender.

Negotiate with a solved ceiling

solve_max_offer turns "is this price OK?" into "here's the most I can pay" — with the binding constraint named, so you know which target you'd trade away first.

claude — offer strategy

> Seller won't budge below $300K. Does the deal still work,
  and where's my ceiling if I need $400/mo?

Called solve_max_offer(strategy: "ltr", targets: {monthlyCF: 400, cocReturn: 0.08}, …)

Ceiling: $303,657 — binding target is your 8% cash-on-cash.
$300K works with your targets intact. You have $3.7K of headroom.

Pick the exit before you buy

One property, six strategies. best_use runs everything the inputs qualify for and ranks them — flip vs BRRRR vs hold stops being a gut call.

claude — flip or rent?

> $180K house, needs $50K of work, ARV $300K, would rent $2,200.
  Flip it, BRRRR it, or hold it?

Called best_use(inputs: {price: 180000, rehabBudget: 50000, arv: 300000, marketRent: 2200, …})

The engine ranks every qualifying strategy and explains the gap —
compare the ranked scores and open each report side by side.

STR underwriting with real market data API KEY

With a key attached, your AI chains str_market_data (nearby listings' ADR, occupancy, revenue quartiles) into analyze_deal's STR strategy — market-grounded assumptions instead of guesses.

What do 2-bedroom STRs actually earn near 30A? Then underwrite this $475K listing at those numbers.
Is my 55% occupancy assumption realistic for this market?

Sanity-check ARVs and rents with comps API KEY

Agents and wholesalers use pull_comps to ground an ARV before running the flip math — real sales, real rents, ZIP market stats.

Pull comps for 4917 Maple Ave, Dallas TX. Is a $295K ARV defensible?
Use the comp-based ARV and re-run the flip. Does it clear the 70% rule?

Builders: BTS vs BTR, decided by the numbers

The newc strategy grades both exits — build-to-sell and build-to-rent — and recommends one with reasons: margin, development spread, capital recovery.

Lot at $85K, 2,400 sqft build at $145/sqft, ARV ~$805K, would rent $3,900. Sell it or hold it?

Building an app instead?

Everything here is also a plain JSON API with the same free tier — see the API reference. The report pages make a zero-UI MVP: analyze via API, hand users the report_url.

Use cases