MarketCheck API & MCP Apps

<img src="https://34682200.delivery.rocketcdn.me/wp-content/uploads/2024/05/cropped-MC-Icon.png.webp" alt="MarketCheck" width="48" align="left" style="margin-right:12px;" />

45 interactive automotive market intelligence dashboards + 7 AI chat demos — usable as MCP UI Apps inside AI assistants, as a standalone web showcase, as embeddable iframe widgets, or as conversational AI chat interfaces.

Powered by MarketCheck real-time automotive data — VIN decoding, ML price predictions, active/sold inventory, and aggregated market analytics. The chat demos showcase the same API capabilities through 7 different chat SDKs.

Live demo: apps.marketcheck.com

---

4 Ways to Use

Mode	Description	Auth Required
Demo	Browse all 45 apps with realistic sample data	None
Live Data	Enter your MarketCheck API key for real market data	API Key (entered in the app UI)
Embed	Embed apps in your website/portal via iframe	OAuth Access Token (secure, 6hr TTL)
MCP / AI	Use inside Claude, VS Code Copilot, Goose, and other MCP hosts	API Key (server-side env var)

Don't have an API key? Sign up free at developers.marketcheck.com

---

Quick Start

Live Demo (hosted)

Visit apps.marketcheck.com — all 45 apps are available in demo mode. Enter your API key in the settings gear to switch to live data.

Self-hosted

# Install dependencies
npm install

# Build everything (gallery + 45 apps + server)
npm run build

# Start the server
npm run serve

Then open:

• Gallery: http://localhost:3001/ — browse all 45 apps
• MCP endpoint: http://localhost:3001/mcp — for AI assistant connectors
• Any app directly: http://localhost:3001/apps/{app-name}/dist/index.html

Demo Mode (no server needed)

Open any built HTML file directly in your browser — all apps have mock data:

open packages/apps/vin-market-report/dist/index.html
open packages/apps/deal-evaluator/dist/index.html
open packages/apps/car-search-app/dist/index.html

---

Connecting as an MCP Server

MCP Server URL

If hosted at apps.marketcheck.com:

https://apps.marketcheck.com/mcp

Authentication for MCP Mode

There are two ways to authenticate:

Option 1: API key in the MCP URL (recommended for individual users)

Pass your MarketCheck API key as a query parameter on the MCP URL:

https://apps.marketcheck.com/mcp?api_key=YOUR_API_KEY

This lets each user provide their own key. No server configuration needed.

Option 2: Server-side environment variable (recommended for shared/hosted deployments)

Set MARKETCHECK_API_KEY on the server. All connected MCP clients share this key automatically — users don't need to provide one.

MARKETCHECK_API_KEY=your_api_key npm run serve

Priority order: If a client provides ?api_key= in the URL, it takes precedence over the server env var for that request. If no URL key is provided, the env var is used as fallback.

Don't have an API key? Sign up free at developers.marketcheck.com

Setting up in Claude

1. Go to Settings → Connectors → Add Custom Connector
2. Enter the MCP server URL:
   • With your own key: https://apps.marketcheck.com/mcp?api_key=YOUR_API_KEY
   • Without key (uses server default): https://apps.marketcheck.com/mcp
3. Start a new chat and ask Claude to use any tool (e.g. "Evaluate this deal: VIN 5TDJSKFC2NS055758")

Custom connectors require a paid Claude plan (Pro, Max, or Team).

Setting up in Claude Code / VS Code / Other MCP Clients

Add to your MCP client configuration:

{
  "mcpServers": {
    "marketcheck-apps": {
      "url": "https://apps.marketcheck.com/mcp?api_key=YOUR_API_KEY"
    }
  }
}

Or without a personal key (server provides its own):

{
  "mcpServers": {
    "marketcheck-apps": {
      "url": "https://apps.marketcheck.com/mcp"
    }
  }
}

How is this different from the MarketCheck MCP Data Server?

MarketCheck offers two separate MCP servers:

Server	Purpose	Tools
MarketCheck MCP Data Server	Raw API access — search, decode, predict, etc.	search_active_cars, decode_vin_neovin, predict_price_with_comparables, etc.
MarketCheck MCP Apps (this project)	Interactive dashboards with visual UI	evaluate-deal, estimate-trade-in, generate-vin-market-report, etc.

Key differences:

• The Data Server returns raw JSON data — the AI model processes and presents it
• The Apps Server returns interactive HTML dashboards that render in the MCP host's UI panel — the user can interact with charts, tables, and filters directly
• The Apps Server calls the same MarketCheck APIs under the hood, but wraps them in rich visual interfaces
• You can use both servers simultaneously — they have different tool names and don't conflict

You do NOT need the Data Server to use the Apps Server. Each is independent. The Apps Server has its own server-side API key and makes its own API calls.

---

All 52 Apps

Consumer (9 apps)

App	Tool Name	Description
VIN Market Report	generate-vin-market-report	Complete VIN-based market intelligence report — embeddable widget
Car Search & Compare	search-cars, compare-cars	Visual car shopping with filters, photo cards, and side-by-side comparison
Car Search	—	Full search with SERP, vehicle details, and natural language search
Deal Evaluator	evaluate-deal	Buy/Negotiate/Pass verdict with gauge, negotiation toolkit, and alternatives
Incentive-Adjusted Deal Evaluator	evaluate-incentive-deal	True out-of-pocket cost after rebates and APR savings
Trade-In Estimator	estimate-trade-in	Instant 3-tier trade-in value with sold comp evidence
Used Car Market Index	get-market-index	Stock-ticker-style dashboard tracking used car prices
OEM Incentives Explorer	oem-incentives-explorer	Cash back, APR, and lease deals by ZIP
Incentive Deal Finder	find-incentive-deals	Search ALL OEM incentives by budget, not by brand

Dealer (5 apps)

App	Tool Name	Description
Lot Pricing Dashboard	scan-lot-pricing	Full lot inventory with market price gaps, aging heatmap, stocking hot list
Stocking Intelligence	stocking-intelligence	Demand heatmap, buy/avoid lists, VIN checker
Pricing Transparency Report	generate-pricing-report	Shareable market report dealers give buyers
Dealer Inventory Fit Scorer	score-dealer-fit	Which cars match your sales DNA? ML-scored acquisitions
Dealer Conquest Analyzer	analyze-dealer-conquest	Find competitors' best-sellers you should stock

Appraiser (4 apps)

App	Tool Name	Description
Appraiser Workbench	appraiser-workbench	Complete vehicle valuation studio
Comparables Explorer	comparables-explorer	Price distribution and market positioning
Depreciation Analyzer	depreciation-analyzer	Track how vehicles lose value over time
Market Trends Dashboard	market-trends-dashboard	The pulse of the automotive market

Dealership Group (3 apps)

App	Tool Name	Description
Group Operations Center	group-operations-center	Every store, one screen
Inventory Balancer	inventory-balancer	Move the right cars to the right stores
Location Benchmarking	location-benchmarking	Rank and compare your locations

Lender (4 apps)

App	Tool Name	Description
Underwriting Decision Support	evaluate-loan-application	Single-loan collateral valuation with LTV forecast
Portfolio Risk Monitor	portfolio-risk-monitor	Track collateral health across your loan book
Lender Portfolio Stress Test	stress-test-portfolio	What-if depreciation scenarios on your loan book
EV Collateral Risk Monitor	ev-collateral-risk	EV vs ICE depreciation risk tracking

Analyst (3 apps)

App	Tool Name	Description
Earnings Signal Dashboard	earnings-signal-dashboard	Pre-earnings channel check for auto tickers
Watchlist Monitor	watchlist-monitor	Morning signal scan across your portfolio
Dealer Group Scorecard	dealer-group-scorecard	Benchmark public dealer groups

Insurer (2 apps)

App	Tool Name	Description
Claims Valuation Workbench	claims-valuation	Total-loss determination with market evidence
Insurance Premium Benchmarker	benchmark-insurance-premiums	Segment-level replacement cost and risk analysis

Manufacturer (2 apps)

App	Tool Name	Description
Brand Command Center	brand-command-center	Your brands vs the competition
Regional Demand Allocator	regional-demand-allocator	Allocate inventory where demand is hottest

Auction House (2 apps)

App	Tool Name	Description
Auction Lane Planner	auction-lane-planner	Plan lanes, price consignments, target buyers
Auction Arbitrage Finder	find-auction-arbitrage	Wholesale vs retail spread — find profit opportunities

Wholesaler (1 app)

App	Tool Name	Description
Wholesale Vehicle Router	route-wholesale-vehicles	Paste VINs, get dealer-match rankings

Cross-Segment (4 apps)

App	Tool Name	Description
EV Market Monitor	ev-market-monitor	The EV transition in one dashboard
VIN History Detective	trace-vin-history	Full listing timeline — dealer hops, price changes, red flags
Market Anomaly Detector	detect-market-anomalies	Find underpriced vehicles and pricing outliers
UK Market Trends	get-uk-market-trends	Macro UK automotive market intelligence

Consumer UK (1 app)

App	Tool Name	Description
UK Market Explorer	search-uk-cars	Search and compare UK car listings in GBP

Dealer UK (1 app)

App	Tool Name	Description
UK Dealer Pricing	scan-uk-lot-pricing	UK lot inventory priced against the market

Auto Media (1 app)

App	Tool Name	Description
Auto Journalist Briefing	generate-market-briefing	One-page market briefing with quotable data points

Fleet Manager (1 app)

App	Tool Name	Description
Fleet Lifecycle Manager	manage-fleet-lifecycle	Fleet values, depreciation, and replacement planning

Rental/Subscription (1 app)

App	Tool Name	Description
Rental Fleet Valuator	value-rental-fleet	Mileage-adjusted fleet valuation with rotation timing

Lender Sales (1 app)

App	Tool Name	Description
Territory Pipeline	territory-pipeline	Find dealers who need floor plan

Chat Demos (7 apps)

Each chat demo uses a different SDK to showcase MarketCheck API capabilities through conversational AI interfaces:

App	SDK	Language	Description
AI Car Advisor	Vercel AI SDK	TypeScript/Next.js	Reference chat with useChat hook, streaming, and tool visualization
Dashboard Copilot	CopilotKit	React/Next.js	AI copilot overlay on existing dashboard UI
MarketCheck Chat	assistant-ui	React/Next.js	Custom-branded noir theme with rich tool result cards
Multi-Platform Bot	Chat SDK	TypeScript	Single codebase deploys to Slack, Discord, Telegram, Teams
Market Analyst	Chainlit	Python	MCP-native chat with built-in tool execution visualization
Quick Market Check	Streamlit	Python	Lightweight chat for data teams
AI Agent Explorer	LangChain	TypeScript/Next.js	LangGraph ReAct agent with visible reasoning chains

Running Chat Demos

TypeScript apps (Vercel AI SDK, CopilotKit, assistant-ui, LangChain):

cd packages/chat/vercel-ai-chat   # or copilotkit-chat, assistant-ui-chat, langchain-agent-chat
cp .env.local.example .env.local  # Add your ANTHROPIC_API_KEY and MARKETCHECK_API_KEY
npm install
npm run dev

Python apps (Chainlit, Streamlit):

cd packages/chat/chainlit-chat    # or streamlit-chat
cp .env.example .env              # Add your keys
pip install -r requirements.txt

# Chainlit:
chainlit run app.py -w

# Streamlit:
streamlit run app.py

Chat SDK Bot (Slack/Discord/Telegram):

cd packages/chat/chat-sdk-bot
cp .env.example .env              # Add API keys + platform bot tokens
npm install
npm run dev

---

API Access Modes

Apps support three data access paths, tried in order:

1. MCP Mode (AI assistants)

When running inside an MCP host (Claude, VS Code, etc.), apps call _safeApp.callServerTool() which routes through the MCP server. The server uses its MARKETCHECK_API_KEY env var — users don't provide a key.

2. Direct API Mode (web/embed)

When loaded in a browser with an API key, apps call the MarketCheck API directly from the browser:

Browser → https://api.marketcheck.com/v2/search/car/active?api_key=KEY&...

No proxy or server needed. The MarketCheck API supports CORS (Access-Control-Allow-Origin: *).

To use: enter your API key in the app's settings gear, or pass it as a URL parameter:

https://apps.marketcheck.com/apps/deal-evaluator/dist/index.html?api_key=YOUR_KEY

3. Demo Mode (no auth)

If no API key is available, apps display realistic mock data. Every app works fully offline.

---

Embedding Apps in Your Portal

Embed any app as an iframe in your website.

Option A: API Key (simple, for internal use)

<iframe
  src="https://apps.marketcheck.com/apps/deal-evaluator/dist/index.html?api_key=YOUR_KEY&embed=true&vin=5TDJSKFC2NS055758"
  width="100%" height="700"
  style="border:none;border-radius:8px;"
></iframe>

Option B: OAuth Token (recommended for production)

Exchange your API key + client secret for a short-lived token (server-side):

curl -X POST https://api.marketcheck.com/oauth2/token \
  -H "Content-Type: application/json" \
  -d '{"grant_type":"client_credentials","client_id":"YOUR_API_KEY","client_secret":"YOUR_SECRET"}'

Then embed with the token:

<iframe
  src="https://apps.marketcheck.com/apps/deal-evaluator/dist/index.html?access_token=TOKEN&embed=true&vin=5TDJSKFC2NS055758"
  width="100%" height="700"
  style="border:none;border-radius:8px;"
></iframe>

URL Parameters

Param	Description
access_token	OAuth token (secure, 6hr TTL)
api_key	API key (for personal/internal use)
embed=true	Hides chrome, full-bleed layout, auto-executes
vin	Pre-populate VIN field
zip	Pre-populate ZIP code
make, model	Pre-populate vehicle selection
miles	Pre-populate mileage
state	Pre-populate state
compact=true	Widget mode (VIN Market Report only)

---

Architecture

marketcheck-api-mcp-apps/
├── package.json                    # Monorepo root (npm workspaces)
├── packages/
│   ├── shared/                     # Shared utilities & types
│   │   └── src/
│   │       ├── types.ts            # Common TypeScript types
│   │       ├── formatters.ts       # Currency, percent, signal classifiers
│   │       ├── index-calculator.ts # Index computation, depreciation, D/S ratio
│   │       └── marketcheck-client.ts # Typed MarketCheck API wrapper (server-side)
│   ├── server/                     # MCP server (44 tools)
│   │   └── src/
│   │       ├── index.ts            # Express + MCP server
│   │       ├── register-app.ts     # Helper to register tool + UI resource pairs
│   │       ├── proxy.ts            # CORS proxy for legacy/fallback mode
│   │       └── tools/              # 44 tool handler files
│   ├── gallery/                    # Web gallery UI
│   ├── apps/                       # 45 app UI folders
│   │   ├── vin-market-report/
│   │   │   ├── src/main.ts         # Full app UI + direct API client
│   │   │   └── dist/index.html     # Built single-file HTML bundle
│   │   └── ... (44 more)
│   └── chat/                       # 7 AI chat demo apps
│       ├── shared/                 # Shared tool definitions & prompts
│       ├── vercel-ai-chat/         # Next.js + Vercel AI SDK
│       ├── copilotkit-chat/        # Next.js + CopilotKit
│       ├── assistant-ui-chat/      # Next.js + assistant-ui
│       ├── chat-sdk-bot/           # Multi-platform bot (Slack/Discord/Telegram)
│       ├── chainlit-chat/          # Python + Chainlit
│       ├── streamlit-chat/         # Python + Streamlit
│       └── langchain-agent-chat/   # Next.js + LangGraph
├── public/                         # Vercel deployment output
├── static/screenshots/             # App screenshots for gallery
└── scripts/                        # Build & migration utilities

How it works

1. The MCP server registers 44 tools, each with a _meta.ui.resourceUri pointing to a ui:// resource
2. When an MCP host (Claude, VS Code, etc.) calls a tool, it also fetches the UI resource — a single-file HTML bundle
3. The app renders in a sandboxed iframe inside the host
4. The app calls app.callServerTool() for data and app.updateModelContext() to push results back to the LLM
5. In web mode, apps call the MarketCheck API directly from the browser (no proxy needed)
6. All apps include mock data fallback for demo mode

Tech stack

• Server: @modelcontextprotocol/sdk + @modelcontextprotocol/ext-apps + Express
• UI: Vanilla TypeScript + Canvas 2D API (no chart libraries)
• Build: Vite + vite-plugin-singlefile → single HTML file per app
• Data: MarketCheck API (12 endpoints: VIN decode, price prediction, active/sold search, listing history, sold summary, OEM incentives, dealer/vehicle ranking, UK markets)

MarketCheck API Endpoints

Endpoint	Path	Purpose
VIN Decode	GET /v2/decode/car/neovin/{vin}/specs	VIN to full vehicle specs
Price Predict	GET /v2/predict/car/us/marketcheck_price/comparables	ML price prediction + comparables
Search Active	GET /v2/search/car/active	Current dealer listings with filters
Search Recent	GET /v2/search/car/recents	Recently sold/expired listings
Car History	GET /v2/history/car/{vin}	Listing timeline for a VIN
Sold Summary	GET /api/v1/sold-vehicles/summary	Aggregated market analytics
OEM Incentives	GET /v2/incentives/by-zip	Manufacturer incentives by ZIP
Rank Dealers	GET /v2/rank/dealers	Dealer-vehicle match scoring
Rank Vehicles	GET /v2/rank/vehicles	Vehicle-dealer fit scoring
UK Active	GET /v2/search/car/uk/active	UK market active listings
UK Recent	GET /v2/search/car/uk/recents	UK market recent listings

---

Deployment

Hosted at apps.marketcheck.com

The production instance is hosted at apps.marketcheck.com.

• Gallery: https://apps.marketcheck.com/
• MCP endpoint: https://apps.marketcheck.com/mcp
• Individual app: https://apps.marketcheck.com/apps/{app-name}/dist/index.html
• Health check: https://apps.marketcheck.com/health

Environment Variables

MARKETCHECK_API_KEY=your_api_key    # Required for MCP mode
PORT=3001                           # Server port (default 3001)

Self-hosting Options

Vercel (recommended for serverless):

npx vercel --prod

Docker / Cloud Run / Railway / Render / Fly.io:

npm install && npm run build
PORT=3001 MARKETCHECK_API_KEY=your_key npm run serve

---

MCP Host Compatibility

MCP Apps are supported by:

• Claude (web) and Claude Desktop
• Claude Code (CLI and VS Code extension)
• VS Code GitHub Copilot
• Goose
• Postman
• MCPJam

---

Development

Build all apps

npm run build

Build a single app

cd packages/apps/deal-evaluator
npx vite build

Add a new app

1. Create packages/apps/my-new-app/ with package.json, tsconfig.json, vite.config.ts, index.html, src/main.ts
2. Create packages/server/src/tools/my-new-app.ts with the tool handler
3. Add to packages/server/src/index.ts toolModules array
4. Add to packages/gallery/src/main.ts APPS array
5. Run npm install && npm run build

Project stats

Metric	Value
Total apps	45 dashboard apps + 7 chat demos
Segments	18 (including Chat Demos)
API endpoints used	12
Built HTML bundles	45 (~400KB each, ~98KB gzipped)
External chart libraries	0 (all Canvas 2D)
Mock data	Every app has full offline fallback

---

License

MIT