Crankly
MCP server for structured public vehicle data from StartMyCar, providing tools to list makes, models, problems, reviews, fuse box data, manuals, guides, and compare models.
README
Crankly
Crankly is a Model Context Protocol (MCP) server for structured public vehicle data from StartMyCar.
What Crankly provides
Crankly exposes the following read-only tools:
| Tool | Description |
|---|---|
listMakes |
Lists available vehicle makes |
listModels |
Lists models for a given make |
getProblems |
Returns reported owner problems |
getReviews |
Returns owner reviews |
getFuseBox |
Returns fuse box entries and circuit references |
getManuals |
Returns owner and service manual links |
getGuides |
Returns guide listings or guide content |
compareModels |
Returns comparison data for two models |
Supported country prefixes:
usaucagbnzzaitfrde
Default country: us
Runtime and package manager support
Crankly runs on Bun.
However, the package itself can be installed or executed using different JavaScript package managers and runners:
- Bun
- npm
- pnpm
- Yarn
Important note
Even if you install Crankly through npm, pnpm, or yarn, the executable still expects Bun to be available on the machine because the server runtime is Bun-based.
In practice this means:
- package manager choice is flexible
- runtime requirement is still Bun
Check Bun:
bun --version
Installation options
1. Run from local source
Use this when developing or testing the repository directly.
git clone https://github.com/Queaxtra/crankly.git
cd crankly
bun install
Then point your MCP client to:
bun /absolute/path/to/crankly/index.ts
2. Run without global installation
This is the easiest published-package flow for one-off or ephemeral use.
Bun
bunx crankly
npm
npx crankly
pnpm
pnpm dlx crankly
Yarn
yarn dlx crankly
These commands still require Bun to be installed because the Crankly executable uses Bun at runtime.
3. Global installation
This is the cleanest option for repeated use across multiple MCP clients.
Bun
bun add -g crankly
npm
npm install -g crankly
pnpm
pnpm add -g crankly
Yarn Classic
yarn global add crankly
After global installation, the executable becomes:
crankly
Verify:
command -v crankly
Quick local validation
Type-check
bun run typecheck
Start from source
bun run index.ts
Inspect with MCP Inspector
If you want to test tools one by one before connecting a client:
bunx @modelcontextprotocol/inspector bun /absolute/path/to/crankly/index.ts
If you already have the executable available globally:
bunx @modelcontextprotocol/inspector crankly
Useful smoke-test inputs:
listMakes
{}
listModels
{
"make": "toyota"
}
getProblems
{
"make": "toyota",
"model": "corolla",
"page": 1
}
getReviews
{
"make": "toyota",
"model": "corolla"
}
getFuseBox
{
"make": "toyota",
"model": "corolla",
"year": 2024
}
getManuals
{
"make": "toyota",
"model": "corolla",
"year": 2022
}
getGuides
{
"make": "toyota",
"model": "corolla"
}
getGuides detail
{
"make": "toyota",
"model": "corolla",
"guide": "tire-pressure"
}
compareModels
{
"make1": "toyota",
"model1": "corolla",
"make2": "honda",
"model2": "civic"
}
MCP client setup
Codex
Codex supports stdio MCP servers and can register them directly from the CLI.
Fastest setup commands
Using Bun runner
codex mcp add crankly -- bunx crankly
Using npm runner
codex mcp add crankly -- npx crankly
Using pnpm runner
codex mcp add crankly -- pnpm dlx crankly
Using Yarn runner
codex mcp add crankly -- yarn dlx crankly
Using global installation
If you already installed Crankly globally:
codex mcp add crankly -- crankly
Using local source directly
codex mcp add crankly -- bun /absolute/path/to/crankly/index.ts
Manual Codex config
Codex commonly uses a TOML config similar to:
Local source
[mcp_servers.crankly]
command = "bun"
args = ["/absolute/path/to/crankly/index.ts"]
startup_timeout_sec = 20
tool_timeout_sec = 120
Bun ephemeral runner
[mcp_servers.crankly]
command = "bunx"
args = ["crankly"]
startup_timeout_sec = 20
tool_timeout_sec = 120
npm ephemeral runner
[mcp_servers.crankly]
command = "npx"
args = ["crankly"]
startup_timeout_sec = 20
tool_timeout_sec = 120
pnpm ephemeral runner
[mcp_servers.crankly]
command = "pnpm"
args = ["dlx", "crankly"]
startup_timeout_sec = 20
tool_timeout_sec = 120
global executable
[mcp_servers.crankly]
command = "crankly"
startup_timeout_sec = 20
tool_timeout_sec = 120
Verify in Codex
codex mcp list
Then try prompts such as:
Use crankly to list the first 10 makes.Use crankly to list Toyota models.Use crankly to get Toyota Corolla reviews.Use crankly to compare Toyota Corolla and Honda Civic.
Claude Code
Claude Code supports stdio MCP servers and can register them from the CLI.
Fastest setup commands
Using Bun runner
claude mcp add --transport stdio crankly -- bunx crankly
Using npm runner
claude mcp add --transport stdio crankly -- npx crankly
Using pnpm runner
claude mcp add --transport stdio crankly -- pnpm dlx crankly
Using Yarn runner
claude mcp add --transport stdio crankly -- yarn dlx crankly
Using global installation
claude mcp add --transport stdio crankly -- crankly
Using local source directly
claude mcp add --transport stdio crankly -- bun /absolute/path/to/crankly/index.ts
Manual Claude Code config
Project-level .mcp.json example:
Local source
{
"mcpServers": {
"crankly": {
"type": "stdio",
"command": "bun",
"args": ["/absolute/path/to/crankly/index.ts"]
}
}
}
Bun runner
{
"mcpServers": {
"crankly": {
"type": "stdio",
"command": "bunx",
"args": ["crankly"]
}
}
}
npm runner
{
"mcpServers": {
"crankly": {
"type": "stdio",
"command": "npx",
"args": ["crankly"]
}
}
}
pnpm runner
{
"mcpServers": {
"crankly": {
"type": "stdio",
"command": "pnpm",
"args": ["dlx", "crankly"]
}
}
}
global executable
{
"mcpServers": {
"crankly": {
"type": "stdio",
"command": "crankly"
}
}
}
Verify in Claude Code
Inside Claude Code:
/mcp
Check that crankly is connected and tools are visible.
Suggested prompts:
Use crankly to list makes.Use crankly to get Toyota Corolla problems page 1.Use crankly to get Toyota Corolla manuals for 2022.Use crankly to get the latest Corolla fuse box data.
OpenCode
OpenCode supports MCP servers through its JSON configuration.
Recommended config locations
~/.config/opencode/opencode.json- project-local
opencode.json
Config examples
Using Bun runner
{
"mcp": {
"crankly": {
"type": "local",
"enabled": true,
"command": ["bunx", "crankly"]
}
}
}
Using npm runner
{
"mcp": {
"crankly": {
"type": "local",
"enabled": true,
"command": ["npx", "crankly"]
}
}
}
Using pnpm runner
{
"mcp": {
"crankly": {
"type": "local",
"enabled": true,
"command": ["pnpm", "dlx", "crankly"]
}
}
}
Using Yarn runner
{
"mcp": {
"crankly": {
"type": "local",
"enabled": true,
"command": ["yarn", "dlx", "crankly"]
}
}
}
Using global installation
{
"mcp": {
"crankly": {
"type": "local",
"enabled": true,
"command": ["crankly"]
}
}
}
Using local source
{
"mcp": {
"crankly": {
"type": "local",
"enabled": true,
"command": [
"bun",
"/absolute/path/to/crankly/index.ts"
]
}
}
}
Verify in OpenCode
Restart OpenCode after updating config.
Suggested prompts:
use crankly and list makesuse crankly and list toyota modelsuse crankly and get toyota corolla reviewsuse crankly and compare corolla with civic
Which setup should you choose?
Best for local development
Use direct source execution:
bun /absolute/path/to/crankly/index.ts
Best for published package without permanent global install
Use one of:
bunx crankly
npx crankly
pnpm dlx crankly
yarn dlx crankly
Best for repeat daily use
Install globally and use:
crankly
Troubleshooting
The command appears to hang
This is expected if you run the server directly.
Example:
crankly
The server is waiting for MCP input. Use an MCP client or MCP Inspector.
The client says the server did not start
Use an absolute path first:
bun /absolute/path/to/crankly/index.ts
This avoids PATH resolution issues.
crankly command is not found
Install globally with your preferred package manager:
bun add -g crankly
or:
npm install -g crankly
or:
pnpm add -g crankly
Then verify:
command -v crankly
npx crankly or pnpm dlx crankly resolves but fails at runtime
Make sure Bun is installed and available in your PATH.
A client does not pick up changes
Restart the MCP client after updating configuration.
License
This project is licensed under the MIT License. See LICENSE.
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.