agentsclimarketplace

Comparative market analysis

Skill realtyapi/realtyapi-skills/skills/comparative-market-analysis

Real estate research skills for AI agents — property briefs, listing search, rentals, CMAs, investment deal screening, and agent finders, powered by RealtyAPI

Install
npx -y skills add realtyapi/realtyapi-skills --skill comparative-market-analysis

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Use when the user wants a CMA — an estimated value range for a property backed by comparable sales and local market trends. Combines the subject property's details, recent comparable/sold homes, the portal valuation, and area market data into a defensible value range with reasoning. Triggers on "run a CMA", "what should this list/sell for", "estimate the value of …", "what are the comps for …", "price this home".

The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

3.6 KB, as published. Nobody here has run it

Comparative Market Analysis (CMA)

Overview

Estimate what a property is worth by triangulating three sources: comparable sales, the portal valuation/estimate, and local market trends. Produce a value range with explicit reasoning — not a single magic number.

Use realtyapi-api for auth and endpoint discovery. This skill builds on property-research (subject details) and adds rigor on comps.

When to Use

  • "What should I list this house for?"
  • "Run a CMA on 123 Main St."
  • "What are the comps and what's it worth?"
  • Pre-listing pricing, offer strategy, or a sanity check on an asking price.

Workflow

  1. Subject property — fetch full details by address/URL (beds, baths, size, lot, year, condition signals). This anchors the comparison.
  2. Comparables — pull comps/similar/nearby sold homes. Filter to genuinely comparable ones: similar size (±~15–20%), type, and recent sales (ideally last 3–6 months) in the same area. Drop outliers and explain why.
  3. Portal valuation — pull the estimate/valuation (zestimate*, estimates).
  4. Market trend — pull area market data (housing_market, market*, housingMarketTrends): direction and $/sqft trend.
  5. Derive the range — base it on comp $/sqft applied to the subject's size, cross-checked against the portal estimate, then nudge for trend and condition. Present a low–high range and a most-likely figure.
  6. Show your work in the report.

Output Format

# CMA: {address}

Subject: {beds}bd/{baths}ba · {size} sqft · {type} · built {year} · {source/link}

## Estimated Value
**{low} – {high}**  (most likely: {point})
- Basis: comp median $/sqft {x} × {size} sqft, adjusted for {trend/condition}
- Portal estimate (cross-check): {value} ({source})

## Comparable Sales
| Address | Sold | Price | Beds/Baths | Size | $/sqft | Dist | Adj. notes |
|---------|------|------:|-----------|-----:|-------:|-----:|-----------|

## Market Context
- Trend: {rising/flat/falling} · Median $/sqft: {} · Days on market: {}

## Confidence & Caveats
- Confidence: {high/med/low} — {# of comps, recency, dispersion}
- Not an appraisal; based on public data and model estimates.

Common Pitfalls

  • Reject bad comps (wrong size/type, stale sales, different sub-market) — quality over quantity. State how many comps survived.
  • Don't blindly trust the portal estimate; use it as a cross-check, not the answer.
  • Always give a range; flag low confidence when comps are thin or scattered.
  • Use sold prices for comps, not active list prices, where possible.
  • This is a CMA, not a licensed appraisal — say so.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.