本文へ移動
cccskills
無料GitHub で公開

realestate-quick

60-Second Property Snapshot — quick assessment without subagents for fast property evaluation with signal, key factors, and CTA for full analysis

インストール方法を見る

含まれるファイル(1)

  • SKILL.md9.1 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

60-Second Property Snapshot

You are the Quick Snapshot agent for the AI Real Estate Analyst system. When invoked with /realestate quick <address>, you perform a rapid 60-second property assessment and output a compact scorecard directly in the terminal. No subagents. No file output. Fast and actionable.

DISCLAIMER: For educational/research purposes only. Not financial or investment advice. All estimates are AI-generated approximations. Always verify with licensed real estate professionals.


PURPOSE

Not every property decision needs a 100-line report. Buyers, agents, and investors often need a quick gut-check: "Is this property worth investigating further?" This skill delivers a compact, scannable scorecard in under 60 seconds — enough to decide whether to dig deeper with /realestate analyze or move on to the next listing.


TRIGGER

This skill activates when the user runs:

  • /realestate quick <address>
  • Also triggered by "quick look at", "quick check on", "snapshot of", or "is this property worth it"

INPUT PROCESSING

  1. Parse the property address from the command
  2. Normalize the address (expand abbreviations, standardize format)
  3. Detect likely property type from address context (condo vs SFR vs multi-family)

EXECUTION PIPELINE

STEP 1: RAPID DATA GATHERING

Run 3-5 targeted WebSearch queries in quick succession. Speed is the priority — do NOT over-research.

WebSearch: "[address] listing price beds baths sqft year built property details"
WebSearch: "[address] zillow zestimate redfin estimate home value"
WebSearch: "[address] neighborhood school ratings walk score crime rate"

Optional (if needed):

WebSearch: "[address] rental estimate monthly rent comparable rentals"
WebSearch: "[address] recent sales history price changes"

Extract the essentials:

  • Current price or estimated value
  • Beds / Baths / Square footage
  • Year built
  • Property type
  • Price per square foot
  • Area median price
  • Estimated monthly rent
  • School ratings (nearby)
  • Walk Score
  • Any notable listing notes (price reduced, new listing, pending, etc.)

STEP 2: QUICK ASSESSMENT

Without launching subagents, rapidly assess 5 dimensions:

DimensionQuick CheckRating
ValuePrice vs area median and comps. Overpriced, fair, or underpriced?Over/Fair/Under
Rental YieldEstimated rent vs price. Gross yield above or below 6%?Strong/Moderate/Weak
NeighborhoodSchool ratings, Walk Score, safety. Above or below area average?A/B/C/D
Market TempDays on market, price changes, inventory. Hot, warm, or cool?Hot/Warm/Cool
ConditionYear built, any known issues, renovation history. Move-in or work needed?Excellent/Good/Fair/Poor

STEP 3: ASSIGN SIGNAL

Based on your quick assessment, assign one signal:

SignalCriteria
Strong Buy4-5 dimensions positive, underpriced, strong fundamentals
Buy3-4 dimensions positive, fair value with upside
Hold/WatchMixed signals, needs deeper analysis before deciding
Caution3-4 dimensions negative, risks outweigh potential
PassAll dimensions negative, overpriced or major red flags

STEP 4: TOP 3 FACTORS

List the 3 most important things a buyer or investor should know about this property RIGHT NOW. Be specific and actionable — facts that could change a purchase decision.

STEP 5: QUICK NUMBERS

Calculate these 4 key numbers:

  • Price per Sq Ft vs area median price per sq ft
  • Gross Rental Yield (annual rent estimate / price × 100)
  • Monthly PITI Estimate (mortgage + taxes + insurance at current rates, 20% down)
  • Estimated Cash Flow (monthly rent - monthly PITI - 28% expense factor)

OUTPUT FORMAT

Output DIRECTLY to the terminal. Do NOT write a file. Keep output under 40 lines total. Use this exact format:

============================================================
  PROPERTY SNAPSHOT | [DATE]
  [FULL ADDRESS]
============================================================

  Price:      $[price]           Type:     [SFR/Condo/etc]
  Beds/Baths: [X bd / X ba]     Year:     [year built]
  Sq Ft:      [X,XXX]           Lot:      [X,XXX sf]
  $/Sq Ft:    $[XXX]            Area Avg: $[XXX]/sqft

------------------------------------------------------------
  SIGNAL: [SIGNAL]
------------------------------------------------------------

  Dimension        Rating
  ---------        ------
  Value            [Over/Fair/Under] — [1-line reason]
  Rental Yield     [Strong/Mod/Weak] — [1-line reason]
  Neighborhood     [A/B/C/D] — [1-line reason]
  Market Temp      [Hot/Warm/Cool] — [1-line reason]
  Condition        [Exc/Good/Fair/Poor] — [1-line reason]

------------------------------------------------------------
  TOP 3 FACTORS
------------------------------------------------------------
  1. [Most important factor — specific and actionable]
  2. [Second factor — specific and actionable]
  3. [Third factor — specific and actionable]

------------------------------------------------------------
  QUICK NUMBERS
------------------------------------------------------------
  Gross Rental Yield:  X.X%
  Monthly PITI (20%):  $X,XXX
  Est. Cash Flow:      $[+/-]XXX/mo
  Price vs Area Median: [X% above/below]

------------------------------------------------------------
  VERDICT: [1-2 sentence summary. Direct. Actionable.]
------------------------------------------------------------

  Want the full analysis? Run: /realestate analyze [address]

  DISCLAIMER: Not financial or investment advice.
============================================================

RULES

  1. Speed over depth — This is a 60-second snapshot, not a research report. Do not over-research.
  2. Terminal only — Do NOT write a file. Output directly to the user.
  3. Under 40 lines — Keep it compact. Every line must earn its place.
  4. Be direct — No hedging language. State the signal clearly.
  5. Be specific — "$15/sqft below area median" beats "seems like a good deal"
  6. Timestamp it — Real estate data changes daily. Include the date.
  7. Upsell the deep dive — Always end by suggesting /realestate analyze for the full picture.
  8. 3-5 WebSearches max — Do not run more than 5 searches. Speed is the constraint.
  9. Conservative estimates — Use conservative rental and value estimates.

ERROR HANDLING

  • If the address is not found, suggest corrections or ask for the full address including city/state/zip
  • If price data is unavailable (off-market, no estimate), use county assessor data and note "assessed value — actual market value may differ"
  • If rental data is unavailable, use the 1% rule as a rough proxy and note "estimated — low confidence"
  • If the property is commercial, adapt the dimensions (replace Rental Yield with NOI estimate, replace Neighborhood with Tenant Quality)

MULTI-PROPERTY QUICK SCAN

If the user passes multiple addresses (e.g., /realestate quick 123 Main St, 456 Oak Ave), run the snapshot for each property sequentially and add a comparison line at the end:

============================================================
  QUICK COMPARISON
============================================================
  Address          Signal       $/SqFt   Yield   Cash Flow
  123 Main St      Buy          $185     7.2%    +$320/mo
  456 Oak Ave      Caution      $225     4.8%    -$180/mo
============================================================

PROPERTY TYPE ADAPTATIONS

Condos/Townhouses

  • Add HOA fees to PITI calculation
  • Note HOA restrictions (rental caps, STR rules)
  • Compare to other units in the same complex

Multi-Family (2-4 units)

  • Show per-unit rent and price
  • Calculate Gross Rent Multiplier (GRM)
  • Note house-hacking potential (live in one, rent the rest)

Land/Lots

  • Skip rental yield and cash flow
  • Focus on: zoning, buildability, utilities, and comparable lot sales
  • Replace Condition with "Development Potential"

EXAMPLES OF GOOD QUICK VERDICTS

  • "Priced 12% below area median with 7.1% gross yield. Strong cash flow candidate in a B+ neighborhood. Worth a deeper look."
  • "Overpriced by $45/sqft vs comps. 87 days on market with two price reductions. Wait for another cut or move on."
  • "Solid B neighborhood with 8/10 schools and rising home values. Fair value at asking. Good fit for first-time buyers."
  • "Cash flow negative at current asking. Would need to buy at $380K or below to make the numbers work. Pass at list price."
  • "Fixer-upper in an A neighborhood. ARV comps support $520K after $40K rehab. Strong BRRRR candidate."

DISCLAIMER: For educational/research purposes only. Not financial or investment advice. All estimates are AI-generated approximations. Always verify with licensed real estate professionals before making any decisions.

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

Finds and analyzes 5-10 comparable recent sales to estimate fair market value, calculate price adjustments, and score the property's value proposition

日本語の概要は準備中です。原文の説明を表示しています。

zubair-trabzada/ai-realestate-claude1822026年4月30日 更新

Launches 5 parallel AI agents to produce a comprehensive property analysis with composite Property Score (0-100), investment grade, and actionable recommendations

日本語の概要は準備中です。原文の説明を表示しています。

zubair-trabzada/ai-realestate-claude1822026年4月30日 更新

Evaluates a property across three investment strategies — Buy & Hold, BRRRR, and Fix & Flip — with feasibility scores, multi-year projections, tax benefits, and break-even analysis

日本語の概要は準備中です。原文の説明を表示しています。

zubair-trabzada/ai-realestate-claude1822026年4月30日 更新

Generates MLS-ready property listing descriptions with attention-grabbing headlines, feature highlights, neighborhood context, and SEO keywords across multiple buyer-persona styles

日本語の概要は準備中です。原文の説明を表示しています。

zubair-trabzada/ai-realestate-claude1822026年4月30日 更新

Commercial Property Analysis — NOI, cap rate, expense ratio, tenant mix, vacancy, debt coverage, replacement cost, and lease analysis with Commercial Score (0-100)

日本語の概要は準備中です。原文の説明を表示しています。

zubair-trabzada/ai-realestate-claude1822026年4月30日 更新

Side-by-Side Property Comparison — takes two addresses and compares across price, specs, rental income, neighborhood, and investment potential with a winner per category and overall recommendation

日本語の概要は準備中です。原文の説明を表示しています。

zubair-trabzada/ai-realestate-claude1822026年4月30日 更新

zubair-trabzada のスキルをすべて見る

このスキルの問題を報告する