Date anchor: Today's date comes from the # currentDate system context. Compute ALL relative dates from it. Example: if today = 2026-03-14, then "prior month" = 2026-02-01 to 2026-02-28, "current month" (most recent complete) = February 2026, "three months ago" = December 2025. Never use training-data dates.

Dealer Intelligence Brief — Prep for a Dealer Sales Call

Profile

Load the marketcheck-profile.md project memory file if exists. Extract: price_range_min, price_range_max, preferred_year_range, max_mileage, approved_makes, approved_segments, ltv_max_pct, lending_type, country, zip, radius. If missing, ask minimum fields. US: search_active_cars, predict_price_with_comparables, get_sold_summary. UK: search_uk_active_cars only (inventory profile only, no LTV analysis). Confirm: "Using profile: [company], [lending_type]". All preference values from profile — do not re-ask.

User Context

Lender sales rep preparing for a dealer meeting. Need to know: what does this dealer sell, how much of it can we finance, what are their pain points (aging, floor plan burden), and what should I say in the meeting.

Gotchas

1. Dealer identification ambiguity. Users may provide a dealer name, domain, or ID. Prefer dealer_id for precision. If only a name is given, use search_active_cars with seller_name to find the dealer_id first, but warn the user if multiple dealers match. If a domain is given, use source=[domain].
2. LTV spot-check is an estimate, not an appraisal. predict_price_with_comparables returns a statistical prediction based on comparables, not a certified valuation. Always label LTV calculations as "estimated" and note the comp count backing each prediction. If a prediction returns fewer than 5 comps, flag it as "low confidence."
3. Floor plan rate varies widely. The $35/day default is an industry midpoint. Actual rates range from $25-$50/day depending on the lender and dealer tier. If the profile includes a floor_plan_rate, use that. Otherwise label as "(estimated at $35/day industry average)."
4. UK profiles — no LTV analysis. predict_price_with_comparables is US-only. For UK dealers, skip steps 4 and 6 (LTV calculations) and note "LTV analysis unavailable for UK market."
5. New vs used inventory separation. Always scope to car_type=used unless the user explicitly asks about new. Mixing new and used produces misleading DOM and pricing metrics.

Workflow: Full Dealer Brief

1. Get full inventory profile — Call mcp__marketcheck__search_active_cars with dealer_id (or source for domain), car_type=used, facets=make|0|20|1,body_type|0|10|1,year|0|10|1, stats=price,dom,miles, rows=0. → Extract only: total_count, avg_price, median_price, avg_dom, avg_miles, make distribution, body_type distribution, year distribution. Discard full response.

2. Get aged inventory — Call mcp__marketcheck__search_active_cars with same dealer, sort_by=dom, sort_order=desc, rows=15. → Extract only: per vehicle — vin, year, make, model, trim, price, miles, dom. Discard full response.

3. Overlay lending criteria — Call mcp__marketcheck__search_active_cars with same dealer, price_range=[min]-[max], year=[year_range], miles_range=0-[max_mileage], rows=0, stats=price. → Extract only: matching_count, avg_price of matching units. Discard full response.

4. LTV spot-check — For the top 5 matching units by price (representative sample), call mcp__marketcheck__predict_price_with_comparables with vin, miles, zip (dealer's zip), dealer_type matching dealer's type. → Extract only: predicted_price per VIN. Calculate estimated LTV = listed_price / predicted_price × 100. Discard full response.

5. Local market context — Call mcp__marketcheck__get_sold_summary with state (dealer's state), inventory_type=Used, ranking_dimensions=make,model, ranking_measure=sold_count, ranking_order=desc, top_n=10, prior month dates. → Extract only: top models by volume. Discard full response.

6. Calculate dealer brief metrics:

   • Lending Fit % = matching_units / total_units × 100
   • Avg LTV = average of spot-check LTVs. Flag units over ltv_max_pct.
   • Floor Plan Burden Estimate:
     • Aged units (DOM > 45) count
     • Estimated burden = aged_units × $35/day × (avg_aged_dom - 45)
   • Aging Pressure:
     • Units > 60 DOM, > 90 DOM, > 120 DOM
     • % of lot over 60 DOM

7. Generate talking points:

   • "You have [X] units in our lending sweet spot ($[min]-$[max], [year range])"
   • "Your average price of $[avg] aligns well with our programs"
   • If aging present: "Your [Y] units over 60 days cost roughly $[Z]/month in floor plan. We can help with competitive terms."
   • If lending_type is floor_plan: "We can floor [X] of your [total] units at [rate]"
   • If lending_type is retail: "We can finance [X] units for your customers with advance rates up to [ltv]%"
   • "The top sellers in your market are [models] — we have strong programs for those"

8. Suggest products:

   • High aging + high inventory → Floor plan financing
   • High fit % + moderate DOM → Retail lending / indirect program
   • Subprime inventory (older, higher miles) → Subprime programs
   • EV inventory → EV-specific lending programs

Output Template

── Dealer Intelligence Brief ── [Month Year] ───────────────

DEALER CARD
Name:           [Dealer Name]
Location:       [City], [State] [ZIP]
Type:           [Franchise/Independent]
Total Used Units: [XX]
Avg Price:      $XX,XXX
Avg DOM:        XX days
Lending Fit:    XX% ([X] of [Y] units qualify)

INVENTORY BREAKDOWN
Make Mix:       [Make1] XX%, [Make2] XX%, [Make3] XX%, Other XX%
Body Types:     [SUV] XX%, [Sedan] XX%, [Truck] XX%, Other XX%
Year Range:     [oldest]-[newest], median [year]
Price Range:    $X,XXX - $XX,XXX, median $XX,XXX

LENDING FIT ANALYSIS
Matching Units: [X] of [Y] (XX%)
Avg Listed Price (matching): $XX,XXX
Estimated Avg LTV:          XX.X% (based on [N]-unit sample)
LTV Distribution:
  Under 100% (under-advanced): X units
  100-110% (standard):         X units
  110-[max]% (elevated):       X units
  Over [max]% (exceeds limit): X units

AGING ANALYSIS
Units > 60 DOM:   [X] ([XX]% of lot)
Units > 90 DOM:   [X]
Units > 120 DOM:  [X]
Est. Floor Plan Burden: $X,XXX/month (aged units, est. $35/day)

TALKING POINTS
1. "[talking point with specific numbers]"
2. "[talking point with specific numbers]"
3. "[talking point with specific numbers]"
4. "[talking point with specific numbers]"
5. "[talking point with specific numbers]"

RECOMMENDED PRODUCTS
- [Product]: [reason based on data]
- [Product]: [reason based on data]

LOCAL MARKET CONTEXT
Top sellers in [state] last month: [Model1], [Model2], [Model3]

Source: MarketCheck market data, [Month Year].

Output

Dealer brief card: Name, City, State, Total Units, Avg Price, Avg DOM, Lending Fit %. Inventory breakdown: make mix, body type mix, year mix. Lending fit analysis: matching units, avg LTV, LTV distribution. Aging analysis: aged units count, floor plan burden estimate. Top 5 talking points for the meeting. Recommended lending products. Local market context (what's selling).

Self-Check (before presenting to user)

1. Lending Fit % math is correct? matching_units / total_units x 100. Verify the numerator comes from the filtered search (step 3) and denominator from the unfiltered search (step 1).
2. LTV calculations use listed_price / predicted_price? LTV = listed_price / predicted_price x 100. Values over 100% mean the dealer is listing above market. Verify this direction — do not invert the ratio.
3. Floor plan burden uses correct units? Burden = aged_count x excess_DOM x daily_rate. "Excess DOM" = avg_aged_dom - aging_threshold (e.g., 45 days). Do not use total DOM — only the portion beyond the threshold.
4. Talking points are specific to THIS dealer? Each point must reference a specific number from the analysis. Generic statements like "we have great rates" add no value.
5. Product recommendations match the data? If recommending floor plan, verify the dealer has significant aging. If recommending subprime, verify the dealer has older/cheaper inventory. Do not recommend products that contradict the inventory profile.
6. Dealer name and location are correct? Cross-check that the dealer_id returned inventory that matches the expected dealer name and city. Mismatched dealer_ids are a common source of embarrassing errors in sales calls.