deal-score

name	deal-score
description	Scores and ranks deal opportunities using a 6-category rubric covering seller motivation, deal economics, financing terms, regional intelligence, prediction market signals, and asset quality. Detects hidden gems — deals that look bad on the surface but are exceptional when regional and macro context is applied. Classifies deals as HOT, WARM, WATCH, or PASS. Use when user says "score deals", "rank opportunities", "/score", "which deals are hot", "find the best deals", "rank what we found", or "score this listing".
allowed-tools	Read, Write, Glob, Grep
metadata	{"author":"Deal Intelligence","version":"2.0.0","category":"workflow-automation","tags":["scoring","ranking","hidden-gems","deal-analysis"]}

Deal Score

Applies the Deal Intelligence scoring rubric to raw or analyzed listings and produces ranked output.

Instructions

Step 1: Load Input

Read processed listings from data/processed/ or raw from data/raw/.

If both exist, prefer processed (already has regionalScore and marketSignal).

Step 2: Apply Base Scoring

For each listing, score each category. See references/scoring-rubric.md for full breakdown.

Category 1 — Seller Motivation (25 pts max)

• DOM: 0-12 pts based on days on market
• Price cuts: 0-6 pts
• Language signals: retiring, estate, must sell, motivated seller
• Check: daysOnMarket, description, priceHistory

Category 2 — Deal Economics (20 pts max)

• Price vs comps: 0-10 pts
• Cap rate (if available): 0-8 pts
• Business price vs NOI: special check for BizBuySell deals
• Auction discount estimate: use starting bid vs estimated value

Category 3 — Financing Terms (15 pts max)

• Owner financing explicit: 10 pts
• Auction (forced sale leverage): 5 pts
• Language suggesting flexibility: 0-5 pts

Category 4 — Regional Intelligence (20 pts max)

• Use regionalScore field if already computed
• If missing: check zone from geos.json, apply zone default score (NCA=65, CO=60, USA=varies)

Category 5 — Prediction Market (10 pts max)

• Use marketSignal field if already computed
• If missing: apply neutral score (5 pts)

Category 6 — Asset Quality (10 pts max)

• Cash flow positive, essential service, value-add potential, scalable

Step 3: Apply Bonuses

Check hidden gem criteria. Any 2+ of these = isHiddenGem: true:

1. Base score under 50 BUT regionalScore 70+
2. Auction/distress price gap vs comps over 30%
3. DOM over 180 + positive market trajectory
4. Business price under 2x verified annual NOI
5. Price cut over 20% + area job growth confirmed
6. Asset looks bad on surface but area fundamentals strong

Apply bonuses:

• Hidden gem: +15
• Auction deal: +10
• Owner finance explicit: +10
• Priority zone (NCA or CO): +5
• 3+ distress signals: +8

Step 4: Classify Tiers

Score	Tier	Action
75+	HOT	Immediate outreach. Pre-write LOI.
55-74	WARM	Deep research. Prepare offer.
35-54	WATCH	Monitor — may ripen.
Under 35	PASS	Archive.

Step 5: Save Scored Output

Save to: data/processed/YYYY-MM-DD_scored.json

Each record must include:

{
  "dealScore": 0,
  "categoryBreakdown": { ... },
  "bonusesApplied": [],
  "isHiddenGem": false,
  "gemReason": "",
  "tier": "HOT|WARM|WATCH|PASS",
  "scoredAt": "ISO-8601"
}

Step 6: Print Summary

Output to console:

• Total scored
• HOT count (and deal names)
• WARM count
• Hidden gems flagged
• File saved to

Examples

Example 1: Score all processed data

User: "/score" Actions: Read data/processed/ → apply rubric to all → classify tiers → save Result: data/processed/YYYY-MM-DD_scored.json

Example 2: Re-score for hidden gems

User: "/hunt hidden-gems" Actions: Load existing scored data → relax DOM threshold to 60 days → re-run gem detection Result: Updated scores, more gems surfaced

Example 3: Score a specific listing

User: "score this listing [URL]" Actions: Fetch listing data → apply full rubric → return score breakdown Result: Score with category breakdown and tier classification

Error Handling

• Missing daysOnMarket: Apply 0 pts for that category, note in breakdown
• Missing askingPrice: Cannot score — flag as "incomplete data"
• Missing regionalScore: Use zone default from geos.json
• Negative cap rate: Flag as "verify NOI" — do not disqualify automatically

References

• references/scoring-rubric.md — Full 6-category scoring breakdown with point tables