vindex-mcp
Provides Canada-aware vehicle intelligence for AI agents, including VIN decoding with factory warranty terms, merged US/Canada recalls, known-issue reports, and itemized purchase costs, with pay-per-call settlement via USDC on Base using x402.
README
vindex-mcp
MCP server for Vindex — Canada-aware, pay-per-call
vehicle intelligence for AI agents. Decode any 17-char VIN with its original
factory-warranty terms, read merged US (NHTSA) + Canada (Transport Canada) recalls,
get LLM-clustered known-issue reports (every issue cites a verified ODI complaint number,
with a bundled reliability-aggregates block), and itemized used-car closing costs for all
of Canada + all 50 US states & DC via one unified purchase-costs endpoint
(country=CA|US). Pay per call with USDC on Base via x402 — no signup, no
bundles, no receipts.
- 3 free tools — no wallet, no API key.
- 4 paid tools — flat per-call USD prices, settled in USDC on Base Mainnet via x402.
Try before you pay: the free get_sample tool returns every paid response shape for one
fixed sample vehicle (2013 Ford F-150).
Quick start
Free tier (no wallet)
{
"mcpServers": {
"vindex": {
"command": "npx",
"args": ["-y", "vindex-mcp"]
}
}
}
You get get_health, get_pricing, get_sample.
Paid tier (with wallet)
Add an EVM wallet private key (0x-prefixed) that holds USDC on Base Mainnet:
{
"mcpServers": {
"vindex": {
"command": "npx",
"args": ["-y", "vindex-mcp"],
"env": { "EVM_PRIVATE_KEY": "0x..." }
}
}
}
The paid tools (decode_vin, get_recalls, get_known_issues, get_purchase_costs)
then settle each call automatically.
Install snippets
Claude Code:
claude mcp add vindex -- npx -y vindex-mcp
# with a wallet:
claude mcp add vindex --env EVM_PRIVATE_KEY=0x... -- npx -y vindex-mcp
Claude Desktop / Cursor: add the JSON block above to your MCP config
(claude_desktop_config.json or .cursor/mcp.json).
Tools
Free tools
| Tool | Description |
|---|---|
get_health |
Vindex API health + configured payment mode (live | mock). |
get_pricing |
Endpoint catalog + per-call USD prices, params, networks, sources (GET /discovery). |
get_sample |
A free sample of any paid response shape (decode / recalls / known-issues / purchase-costs) for the fixed sample vehicle. decode includes the factory-warranty block; known-issues includes the bundled reliability block; purchase-costs returns both CA + US. |
Paid tools
| Tool | Endpoint | Price (USD) |
|---|---|---|
decode_vin |
GET /v1/decode |
$0.01 — normalized VIN decode + original factory-warranty terms (basic/powertrain/corrosion/roadside/EV-battery). |
get_recalls |
GET /v1/recalls |
$0.01 — merged US + Canada recalls + full decoded vehicle. |
get_known_issues |
GET /v1/known-issues |
$0.05 — LLM-clustered failure modes + bundled reliability block + full decode + complaintsAnalyzed. |
get_purchase_costs |
GET /v1/purchase-costs?country=CA|US |
$0.02 — unified itemized closing costs for Canada (provinces/territories) or the US (states + DC). |
Prices are flat per call — no bundles, no entitlements. Confirm live prices any time
with get_pricing. Purchase-costs is ONE endpoint selected by country=CA|US: CA requires
province (optional family_gift, buyer_has_plates), US requires state (optional
trade_in, local_rate).
How payments work (x402)
Vindex speaks the x402 exact scheme. The API's 402 advertises a
Base (USDC, EIP-3009) rail; this MCP settles on Base:
- The MCP requests a paid endpoint. With no payment, the API replies HTTP 402 with the accepted terms.
- The MCP settles the call — via the standard
@x402/fetchV2 client on a live API, or the mock header when the API is inmode: mock— and retries. - The API returns the data plus a
paymentblock (txHash,network,priceUsd,payer).
Compute-first, settle-after: Vindex computes the full answer before charging, so
you are never billed for an error, a validation failure, or a below-threshold answer
(e.g. get_known_issues refuses, uncharged, below 15 complaints).
Mock vs live
get_health reports mode. When the API is in mock mode the paid tools work without
a wallet (a synthetic payer is used and no chain settlement occurs) — useful for trying
the full flow for free. In live mode the paid tools require EVM_PRIVATE_KEY with
USDC on Base Mainnet; a paid call without a key returns a clear error.
Environment variables
| Variable | Required | Description |
|---|---|---|
EVM_PRIVATE_KEY |
No | 0x-prefixed wallet key with USDC on Base Mainnet. Required to pay a live API. Free tools work without it; against a mock API the paid tools also work without it. |
VINDEX_API_URL |
No | Override the API base URL (default https://api.vindexapi.dev). |
Disclaimer
Vindex provides informational vehicle data aggregated from public sources (NHTSA, Transport Canada). It is not professional, legal, or purchasing advice, and is provided as-is without warranty. Recall/complaint data reflects the source databases at fetch time. Contains information licensed under the Open Government Licence – Canada. See https://api.vindexapi.dev/terms.txt.
Docs for agents: https://api.vindexapi.dev/llms.txt · full: https://api.vindexapi.dev/llms-full.txt
License
MIT © jcislo
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.