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.

Pitch Deck Data — Market Intelligence for Dealer Meetings

Profile

Load the marketcheck-profile.md project memory file if exists. Extract: lending_type, target_states, price_range_min, price_range_max, country. If missing, ask for state. US: get_sold_summary, search_active_cars. UK: search_uk_active_cars only (supply data only — limited talking points). Confirm: "Using profile: [company], [lending_type]". All preference values from profile — do not re-ask.

User Context

Lender sales rep preparing data points for a dealer meeting. The goal is to arm the rep with local market intelligence they can cite during the conversation to build credibility and frame their lending products as informed, data-backed solutions.

Gotchas

1. US-only for sold data. get_sold_summary is US-only. For UK profiles, only supply-side talking points are available (active inventory counts, DOM, pricing from search_uk_active_cars). Clearly note that volume trend and velocity data are unavailable for UK.
2. State vs ZIP scoping. Talking points scoped to a state are broad. If the dealer operates in a metro area, consider using zip + radius=50 for search_active_cars supply calls to make the data more locally relevant. Use state-level for get_sold_summary (which only supports state).
3. Do not present raw API numbers as precision. Round sold counts to nearest 100 for state-level data, nearest 10 for local. Round prices to nearest $100. Dealers will question exact numbers that imply false precision.
4. Stale month risk. If today is the 1st-5th of the month, prior month sold data may be incomplete. Use the month before that as the primary period, and note "(data may still be settling for [recent month])."
5. Never reveal lender-internal metrics. Talking points are for external use with dealers. Do not include penetration rates, spread assumptions, or internal scoring in the output.

Workflow: Local Market Talking Points

1. What's selling — Call mcp__marketcheck__get_sold_summary with state, inventory_type=Used, ranking_dimensions=make,model, ranking_measure=sold_count, ranking_order=desc, top_n=10, date_from (first of prior month), date_to (last of prior month). → Extract only: top 10 models with sold_count, average_sale_price. Discard full response.

2. Volume trend — Same call with date range shifted back one month. Compare total volume. → Extract only: total sold_count. Discard full response.

3. Pricing trends — Call mcp__marketcheck__get_sold_summary with state, inventory_type=Used, ranking_dimensions=body_type, ranking_measure=average_sale_price, ranking_order=desc, top_n=8, current + 3 months ago. → Extract only: per body_type — average_sale_price for both periods. Discard full response.

4. Supply snapshot — Call mcp__marketcheck__search_active_cars with state, car_type=used, seller_type=dealer, stats=price,dom, rows=0. → Extract only: total active, median_price, avg_dom. Discard full response.

5. Calculate days supply — active_inventory / (monthly_sold / 30). Under 45 = tight (dealers stocking aggressively), 45-75 = balanced, over 75 = building (dealers may slow purchases).

6. Generate talking points — Format as quotable statements:

   • "The [state] used car market moved [X] units last month, [up/down Y%] from the prior month."
   • "Average used vehicle transaction price is $[Z], [trending up/down/flat]."
   • "[Segment] is the fastest-turning category at [X] days on market."
   • "Top sellers right now are [Model 1], [Model 2], and [Model 3] — we have strong programs for all of these."
   • "Current inventory levels sit at [X] days supply — dealers [are stocking aggressively / have room to grow / may be cautious]."
   • "SUV transaction prices [rose/fell] [X]% over 3 months — [strong/weakening] residuals for your lending portfolio."
   • If lending_type is floor_plan: "Average DOM across the state is [X] days — every day above 30 is costing dealers money."

7. Tailor to specific dealer (optional) — If user specifies a dealer, add:

   • "[Dealer] has [X] units in our lending sweet spot"
   • "Their average DOM is [X] — [above/below] the state average"

Output Template

── Pitch Data: [State] Market Intelligence ── [Month Year] ──

MARKET OVERVIEW
- "The [state] used car market moved ~[X] units last month, [up/down ~Y%] from [prior month]."
- "[Segment] is the fastest-turning category at [X] days on market."
- "Current inventory levels sit at [X] days supply — dealers [interpretation]."
- "Top sellers right now are [Model 1], [Model 2], and [Model 3]."

PRICING TRENDS
- "Average used vehicle transaction price is $[Z], [trending up/down/flat] over 3 months."
- "SUV transaction prices [rose/fell] [X]% over 3 months — [strong/weakening] residuals."
- "[Body type] pricing spread: $[low]-$[high], median $[mid]."

OPPORTUNITY FRAMING
- "We have strong programs for [top-selling models] — competitive rates for your buyers."
- "Every day above 30 DOM costs dealers money — our floor plan rates start at [X]."
- "[If specific dealer]: You have [X] units in our lending sweet spot ($[min]-$[max])."

Source: MarketCheck market data, [Month Year], [State].

Output

Formatted talking points section with specific numbers. Each point on its own line, ready to copy-paste. Include source citation: "Source: MarketCheck market data, [Month Year], [State]." Group into: MARKET OVERVIEW (3-4 points), PRICING TRENDS (2-3 points), OPPORTUNITY FRAMING (2-3 points). Total ~10 points maximum — concise enough for a sales conversation.

Self-Check (before presenting to user)

1. All numbers are rounded appropriately? State-level volumes rounded to nearest 100, local to nearest 10, prices to nearest $100. No false precision.
2. Talking points are dealer-facing? No internal lender metrics (penetration, spread, scoring) in the output. Everything should be something the rep can say out loud.
3. Trend direction matches the data? If volume went from 5,200 to 5,100, that is "roughly flat" or "down ~2%", not "declining sharply." Verify directional language matches magnitude.
4. Days supply calculation is correct? days_supply = active_inventory / (monthly_sold / 30). Under 45 = tight, 45-75 = balanced, over 75 = building. Verify the interpretation matches the number.
5. Source citation is present? Every output must end with "Source: MarketCheck market data, [Month Year], [State]."
6. No more than 10 points total? The rep needs a quick cheat sheet, not a research paper. If you generated more than 10, prune the weakest.