/weekly-review

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.

Weekly dealer review using parallel sub-agents. Per-location or group rollup.

Step 1: Verify dealer group profile

Read the marketcheck-profile.md project memory file. Parse the JSON content after the --- frontmatter. Missing -> "Run /onboarding first." Stop. Extract group_name, locations[], preferences.

Speed rule — profile-read-once: Pass the extracted profile fields (dealer_id, source, country, zip/postcode, state/region, radius, aging_threshold, dealer_type, franchise_brands) directly to all sub-agents in their prompt. Sub-agents should NOT re-read the profile.

Step 2: Determine scope

$ARGUMENTS: location name -> match. "all" or empty -> ask. Extract per-location fields. dealer_id null -> skip. Tool routing: US = all agents. UK = lot-scanner only.

Step 3: Per-location Wave 1 -- Two agents in parallel

Agent A: lot-scanner -- Spawn dealership-group:lot-scanner: complete used inventory, sort_by=dom desc. Paginate ALL.

Agent B: market-demand-agent (US only) -- Spawn dealership-group:market-demand-agent: hot list + demand snapshot for state, dealer_type, zip, radius. Previous month. Sections: hot_list, demand_snapshot.

Step 4: Per-location Wave 2 -- Price all inventory

After lot-scanner: Agent C: lot-pricer (US only) -- price ALL vehicles. UK: Comp medians from search_uk_active_cars.

Step 5: Per-location report

Section 1 (Lot Scan): pricing table sorted by most overpriced — one row per unique VIN (lot-scanner already deduped; show ×N sites if cross-listed). Section 2 (Hot List): cross-ref with inventory. Section 3 (Demand Snapshot). Price Reductions: deduplicate by VIN, show largest reduction, note ×N sites if cross-listed. Top 5 actions by $ impact. UK: Sections 2-3 replaced with "Requires US sold data."

Step 6: Group weekly rollup (if "all")

Per-location stats MUST be queried individually. For each location, call search_active_cars(source=<location_domain>, car_type="used", rows=1, stats="price,miles,dom") and use that location's own stats object. Never copy group-level stats across rows — each location has a distinct pricing and DOM profile.

Table: location, units, avg_price (per-location), avg_dom (per-location), overpriced/at-market/underpriced counts, hot list match, stocking gaps. Group stocking priorities (most needed models + best location). Cross-location transfer opportunities.