propertyiq — Small Residential Investment Underwriter

Slash Command

This skill can be invoked directly with: /propertyiq [listing URL] /propertyiq [URL1] [URL2] [URL3] /propertyiq [address] [asking price] /propertyiq (no argument — prompts user for URL or address)

Accepts 1, 2, or 3 listing URLs in a single invocation. When multiple URLs are provided, run a full individual report for each property, then generate a side-by-side comparison. When called via slash command, skip the ambient trigger detection and proceed directly to Step 1 with whatever the user provided after the command.

Scope

propertyiq covers small residential investment properties only:

Single-Family Rentals (SFR) — 1 unit
Small Multifamily — Duplex, Triplex, Quadplex (2–4 units)
Mid Multifamily — 5–20 units

Out of scope (not handled by this skill):

Large multifamily (21+ units) — requires agency debt underwriting, T-12 actuals, loss-to-lease
Mobile home parks — land-only income model, pad rent structure
Short-term rentals (Airbnb/VRBO) — requires ADR/occupancy modeling, platform fees
Student housing — per-bed pricing, academic-year lease structure
Commercial / mixed-use — entirely different income and expense framework

If a user provides a property type that is out of scope, tell them clearly and explain what type of analysis would be needed instead.

Step 1 — Detect Mode & Scrape Listings

1a. Detect Single vs. Multi-Property Mode

Count the number of listing URLs in the user's message:

1 URL → Single property mode. Run Steps 2–6 once. Output one .md report file.
2–3 URLs → Multi-property mode. Run Steps 2–5 for EACH property independently. Then run Step 6 (Comparison Report) after all individual reports are complete. Output one .md file per property PLUS one comparison .md file.

Maximum supported: 3 URLs. If the user pastes more than 3, process the first 3 and note that only 3 are supported per run.

1b. Scrape Each Listing

For each URL, use web_fetch to extract:

Property facts:

Address, city, state, zip
Property type (SFR, duplex, triplex, quadplex, 5–20 unit, mixed-use)
Number of units
Year built
Square footage (total and per unit if available)
Bedrooms/bathrooms per unit
Asking price
Listed rents (if shown) — current and/or pro-forma
Listed expenses (taxes, insurance, HOA if any)
Utilities structure (tenant-paid vs. owner-paid — list which ones)
Parking, laundry, amenities notes
Days on market, price history if visible
Any mentioned cap rate or GRM from the listing

If the page is behind a login or returns incomplete data, use web_search to look up the address directly (e.g. "123 Main St Springfield IL" site:zillow.com OR site:realtor.com) and try web_fetch on the best result.

Step 2 — Fill Gaps with Market Research

2a. Rent Comps (ALWAYS run this if listing rents are missing or unverified)

Search for comparable rents in the same zip or neighborhood:

web_search: "[city] [state] [bed/bath] rental price [year]"
web_search: "[zip code] average rent [unit type]"
web_search: "rentometer [city] [state] [bedroom count]"

Pull 2–3 data points. Use the median of comps as your rent estimate. Note the range. If the listing already shows rents, still search comps to validate — flag a delta >15% as a risk.

2b. Local Property Tax

If not on the listing, search: "[address] property tax" or "[county] property tax rate". Use assessed value × mill rate if you find it. Estimate conservatively if uncertain.

2c. Insurance Estimate

If not listed, use regional benchmarks from references/insurance-benchmarks.md.

2d. Safety Index — Crime & Neighborhood Data

ALWAYS run this step for every analysis. Search for crime and safety data for the property's zip code and neighborhood. This data directly affects vacancy risk, tenant quality, insurance premiums, and long-term appreciation — all of which are investor concerns.

Search queries to run:

web_search: "[zip code] crime rate statistics [year]"
web_search: "[city] [neighborhood] crime index safety score"
web_search: "NeighborhoodScout [zip code] crime"
web_search: "SpotCrime OR CrimeGrade [zip code]"

Data points to collect (use what's available — not all will be found):

Overall crime index / safety score (e.g. CrimeGrade letter grade A–F, NeighborhoodScout percentile, SpotCrime score)
Violent crime rate (per 1,000 residents) vs. national average
Property crime rate (per 1,000 residents) vs. national average — most relevant for landlords
Trend: is crime increasing, decreasing, or stable over last 3 years?
Any neighborhood-specific context (proximity to high-crime areas, gentrification, etc.)

Grading scale to apply consistently: A (Very Safe): Crime well below national avg — top 20% safest nationally B (Safe): Crime meaningfully below national avg C (Average): Near national avg — acceptable but flag for investor awareness D (Below Average):Crime above national avg — elevated vacancy/tenant/insurance risk F (High Risk): Crime significantly above national avg — major investment risk flag

Investor impact framing — always include at least one of these observations:

High property crime → elevated insurance premiums, higher vacancy, tenant turnover risk
High violent crime → impacts tenant demand, lender scrutiny, appreciation drag
Improving trend (crime declining) → positive for appreciation; potential value-add signal if buying ahead of neighborhood improvement; note gentrification may displace existing tenants
Worsening trend (crime rising) → increasing vacancy risk, insurance pressure, value drag

If data is unavailable: Note "Safety data unavailable for this zip — recommend checking CrimeGrade.org, NeighborhoodScout.com, or local PD crime map before proceeding."

Step 3 — Confirm Assumptions with User

Single property mode: Present scraped data and proposed assumptions for the one property.

Multi-property mode: Present a combined confirmation block showing all properties side-by-side. Use the same financing defaults for all properties unless the user specifies otherwise (different property types may have different defaults per the reference table).

Example (single property):

I found the following from the listing. Please confirm or correct before I run the analysis:

  Address:          123 Main St, Trenton, NJ 08611
  Property Type:    Triplex (3 units)
  Year Built:       1922
  Asking Price:     $485,000
  Unit Mix:         3x 2BR/1BA (~750 sqft each)
  Current Rents:    $1,400/unit/mo (listing)
  Rent Comps:       $1,350–$1,550/mo (Zillow/Rentometer median ~$1,450)
  Property Tax:     $9,200/yr (public record)
  Utilities:        Tenant-paid (all)

  Financing Defaults I'll use:
    Down Payment:     25% ($121,250)
    Interest Rate:    7.25% (current 30yr commercial avg)
    Amortization:     25 years
    Vacancy:          7% (multifamily standard)
    CapEx Reserve:    $125/unit/mo ($4,500/yr)
    Maintenance:      $60/unit/mo ($2,160/yr)
    Mgmt Fee:         8% of EGI (standard 3rd-party)
    Points:           0 (conventional loan)

  Override anything? Or type "run it" to proceed.

Example (multi-property — condensed):

Found 3 properties. Same financing defaults will apply to all unless you specify otherwise.

  Property 1:  123 Main St, Trenton NJ  — Triplex — $485,000 — Rents $1,400/unit
  Property 2:  456 Oak Ave, Newark NJ   — Duplex  — $320,000 — Rents $1,600/unit
  Property 3:  789 Pine St, Camden NJ   — SFR     — $185,000 — Rent  $1,800/mo

  Financing defaults: 25% down | 7.25% | 25yr amort | market vacancy | standard reserves

  Override anything? Or type "run it" to proceed.

If the user says "run it" or provides no corrections, proceed with defaults.

Step 4 — Run the Underwriting Calculations

Industry-Standard Defaults by Property Type

Read references/defaults-by-property-type.md for the full table. Summary:

Metric	SFR	2–4 Unit	5–20 Unit
Down Payment	25%	25%	25–30%
Interest Rate	current 30yr conv.	current 30yr conv.	current commercial avg
Amortization	30 yr	25 yr	25 yr
Vacancy	5%	7%	8%
CapEx Reserve	5–8% gross rent	$100–150/unit/mo	$75–100/unit/mo
Maintenance	$75–125/mo	$50–75/unit/mo	$40–60/unit/mo
Mgmt Fee	10% EGI	10% EGI	8% EGI
OpEx Ratio Target	45–55%	45–55%	45–55%

Always use the midpoint or conservative end of ranges — these are feasibility screens, not best-case projections.

Calculations to Perform

Gross Potential Income (GPI)      = sum of all unit monthly rents × 12
Vacancy Loss                      = GPI × vacancy rate
Effective Gross Income (EGI)      = GPI − Vacancy Loss
Operating Expenses                = Taxes + Insurance + Mgmt Fee + CapEx Reserve
                                    + Maintenance/Repairs + Other
  NOTE: CapEx Reserve and Maintenance are SEPARATE line items:
    CapEx Reserve   = major capital items (roof, HVAC, appliances) — set aside, not spent annually
                      SFR: 5–8% gross rent | 2–4 unit: $100–150/unit/mo | 5–20 unit: $75–100/unit/mo
    Maintenance     = ongoing minor repairs
                      SFR: $75–125/mo | 2–4 unit: $50–75/unit/mo | 5–20 unit: $40–60/unit/mo
Net Operating Income (NOI)        = EGI − Operating Expenses
Annual Debt Service               = monthly P&I × 12  [use standard amortization formula]
Annual Cash Flow                  = NOI − Annual Debt Service
Monthly Cash Flow                 = Annual Cash Flow / 12

Cap Rate                          = NOI / Purchase Price
Cash-on-Cash Return (CoC)         = Annual Cash Flow / Total Cash Invested
  Show TWO versions:
    CoC (Down Payment only)       = Annual Cash Flow / Down Payment
    CoC (All-In)                  = Annual Cash Flow / (Down Payment + Closing Costs + Points)
  Primary CoC = down payment only (matches investor convention).
  Total Cash to Close = Down Payment + Closing Costs + Origination Points
    Closing Costs = 2.5% (SFR/2–4 unit) or 3% (5–20 unit) of purchase price
    Origination Points = 0 for conventional | 1–2% of loan amount for commercial/DSCR loans
Debt Service Coverage Ratio (DSCR)= NOI / Annual Debt Service
Loan-to-Value (LTV)               = Loan Amount / Purchase Price
Gross Rent Multiplier (GRM)       = Purchase Price / GPI
Break-Even Rent/Unit              = (Annual Debt Service + Operating Expenses) / (units × 12)
Price Per Unit                    = Purchase Price / units
1% Rule Check                     = (Monthly Rent / Purchase Price) × 100
  ≥1.0%  → passes quick screen; worth full underwriting
  0.7–1.0% → borderline; depends on market and appreciation thesis
  <0.7%  → fails rule; cash flow very unlikely at conventional financing
  Note: 1% rule is a blunt filter, not a decision — always run full NOI regardless

For monthly P&I use: M = P × [r(1+r)^n] / [(1+r)^n − 1] where P = loan amount, r = monthly rate, n = amortization months

Step 5 — Build the Individual Property Report (.md)

Output each report as a properly formatted Markdown file using the template below. Use Markdown headers, tables, and bold — NOT ASCII dividers. Save as:

Single property: propertyiq-[short-address].md (e.g. propertyiq-403-leverett-staten-island.md)
Multi-property: propertyiq-[short-address]-1.md, propertyiq-[short-address]-2.md, etc.

Read templates/report.md and fill every [placeholder] with the calculated values for this property. Do not skip any section. Do not change the structure.

Step 6 — Build the Comparison Report (.md) [MULTI-PROPERTY MODE ONLY]

After all individual reports are complete, generate a single comparison file named propertyiq-comparison.md. This file surfaces the key decision-making metrics side-by-side so the investor can rank and choose.

Read templates/comparison.md and fill every [placeholder] with the calculated values for each property. Do not skip any section. Do not change the structure.

Step 7 — Data Quality Notes

At the end of each individual report, be transparent about confidence level:

HIGH: All key inputs scraped or confirmed from public records
MEDIUM: 1–2 inputs estimated (flag which)
LOW: Rents or expenses are estimated — flag prominently and recommend user verification

References

references/defaults-by-property-type.md — Full default assumptions table by property type
references/insurance-benchmarks.md — Regional insurance cost estimates
references/cap-rate-benchmarks.md — Market cap rate ranges by metro tier

Read these reference files when you need the detailed benchmarks. They are not loaded by default to keep context lean — pull them as needed.

propertyiq

What is propertyiq?

Ask in your favorite AI

Documentation