Insurer PluginHeavySkillv0.12.3

Vehicle Appraiser

Insurance valuation with comparable evidence — total-loss threshold, settlement range, and CPO premium built in

appraise this vehicleinsurance valuationcomparable analysisfair market valuepre-loss valueappraisal report
Heavy
Complexity
5–7
API Calls
$0.005 – $0.014
MC API Cost
$0.08 – $0.22
Total Cost

Purpose

Insurance adjusters need a valuation that does two jobs simultaneously: establish the pre-loss fair market value with enough comparable evidence to withstand dispute, and apply the total-loss threshold to determine whether a repair decision or a settlement decision is warranted. Book values fail at the first job; manual comp pulls fail at the second because they lack the algorithmic prediction to anchor the range.

Vehicle Appraiser for insurers runs the full three-source valuation (algorithmic prediction + active comparables + sold transactions) against a 100-mile radius to ensure sufficient comparable evidence for dispute resolution. The total-loss threshold is automatically calculated from the condition-adjusted FMV and the insurer's configured threshold percentage. CPO vehicles get separate certified and non-certified valuations with the premium quantified. Regional price variance is available when settlement disputes involve geographic pricing arguments.

Audience:Insurance adjustersTotal-loss specialistsClaims managersUnderwriters validating coverage amounts

How It Works

Execution flow. MCP tool calls are shown inline on each step.

01
Profile Load

Reads marketcheck-profile.md for ZIP, state, total_loss_threshold_pct, and default_comp_radius. US-only. Extracts role (adjuster/underwriter/claims_manager) to frame output appropriately.

02
VIN Decodedecode_vin_neovin

Calls decode_vin_neovin to lock year, make, model, trim, body type, drivetrain, engine, and transmission. Assumed trims lose credibility in appraisal disputes — the decode is mandatory for defensible insurance valuations.

03
Dual Price Predictionpredict_price_with_comparables

Calls predict_price_with_comparables with dealer_type=franchise (retail replacement cost) and dealer_type=independent (wholesale proxy). For CPO vehicles: additional call with is_certified=true to quantify the certified premium. Settlement anchor = franchise retail replacement cost.

↔ Parallel Execution

04a
Active Comparable Pullsearch_active_cars

Calls search_active_cars with YMMT, ZIP, radius=100mi (wider than trade appraisals — insurance disputes require broader comparable evidence), mileage band ±15K, sorted by price ascending, rows=20.

04b
Sold Transaction Pullsearch_past_90_days

Parallel call to search_past_90_days with the same YMMT and location filters. Sold transactions are the most defensible evidence type — they reflect completed market transactions, not asking prices.

05
FMV Synthesis

Combines the three data sources into a condition-adjusted FMV range (low/mid/high). Condition adjustment: Clean = high end of comp range, Average = midpoint, Rough = low end minus condition discount.

06
Total-Loss Threshold

Repair cost threshold = FMV × total_loss_threshold_pct. If repair cost was provided, outputs TOTAL LOSS or NOT TOTAL LOSS determination. If not provided, outputs the threshold figure for the adjuster to apply against the actual repair estimate.

07
Insurance Report Output

Delivers: vehicle ID summary, valuation table (franchise/independent/condition-adjusted FMV/comp ranges/confidence), total-loss threshold and determination, settlement range (low/mid/high), comparable data tables, and methodology notes with condition adjustments and caveats.

MCP Tool Calls

ToolCallsPurpose
decode_vin_neovin1Exact spec confirmation — mandatory for dispute-defensible valuations
predict_price_with_comparables2–3Franchise retail and independent wholesale predictions; third call for CPO vehicles
search_active_cars1–2Active comparables within 100mi; second call for CPO comparables if applicable
search_past_90_days1Sold transaction evidence for the last 90 days
get_car_history1 (optional)Historical listing trajectory for pre-loss value documentation

Example Output

INSURANCE VALUATION REPORT
════════════════════════════════════════════════
Vehicle:    2021 Ford F-150 XLT 4WD SuperCrew
VIN:        1FTFW1E53MFC12345
Odometer:   44,200 miles | Pre-loss condition: Clean
Purpose:    Total-loss determination | Claims
Location:   Denver, CO (ZIP 80201) | Radius: 100 mi

VALUATION SUMMARY
  Franchise (Retail) Predicted:     $38,200  (based on 19 comps)
  Independent (Wholesale) Predicted: $34,100  (based on 12 comps)
  Active Comparable Range:           $35,100$43,400  (24 comps)
  Sold Transaction Range:            $33,400$40,900  (16 sales, 90 days)
  ──────────────────────────────────────────────────────
  Pre-Loss FMV (Clean condition):   $38,200  (franchise retail — replacement cost)
  Settlement Range — Low:           $35,100  (25th pct sold)
  Settlement Range — Mid:           $38,200  (FMV)
  Settlement Range — High:          $41,800  (75th pct sold)
  Confidence:                       HIGH (24 active comps, 16 sold)

TOTAL-LOSS THRESHOLD
  Pre-Loss FMV:           $38,200
  Threshold (75%):        $28,650
  ──────────────────────────────────────────────────────
  This vehicle is a total loss if repair costs exceed $28,650.

  Estimated repair cost:  $31,400⚠ TOTAL LOSS CONFIRMED (+$2,750 above threshold)

Cost Estimate

TierHeavy
API Calls5–7
MC API — best$0.005
MC API — worst$0.014
Claude — best$0.07
Claude — worst$0.20
Total range$0.08$0.22

300 claims appraisals/month ≈ $24–66

Limitations

  • US-only — VIN decode, ML prediction, and sold history are US-only.
  • 100-mile comp radius is wider than trade appraisals — by design, to ensure sufficient dispute-defensible evidence. May be expanded to 150-200 miles for rural markets.
  • CPO detection requires either VIN history pull or user confirmation; adds one additional API call.
  • The 75% total-loss threshold default may differ from state-specific or insurer-specific thresholds — always verify against the profile or state regulations.
Insurer Plugin
View all skills & commands