MCP (Model Context Protocol) server for Malaysia's public transit system, providing real-time bus and train information across 10+ cities in Malaysia.
MCP Endpoint: https://mcp.techmavie.digital/malaysiatransit/mcp
Analytics Dashboard: https://mcp.techmavie.digital/malaysiatransit/analytics/dashboard
Data Source: Malaysia Transit Middleware
- Features
- Architecture
- Quick Start
- Installation
- Configuration
- Available Tools
- Usage Examples
- AI Integration Guide
- Supported Service Areas
- Deployment
- Troubleshooting
- Contributing
- 11 Operational Service Areas + 1 Under Maintenance + 1 Coming Soon across Malaysia
- Klang Valley (Rapid Rail KL, Rapid Bus KL, MRT Feeder)
- Penang (Rapid Penang, Penang Ferry, KTM Komuter Utara)
- Ipoh (BAS.MY Ipoh, KTM Komuter Utara)
- Seremban (BAS.MY Seremban, KTM Intercity)
- Kangar (BAS.MY Kangar, KTM Komuter Utara)
- Alor Setar (BAS.MY Alor Setar, KTM Komuter Utara)
- Kota Bharu (BAS.MY Kota Bharu, KTM Intercity)
- Kuala Terengganu, Melaka, Kuching (BAS.MY only)
- Johor (BAS.MY Johor Bahru + Bas Muafakat Johor — 41 BMJ routes via PAJ adapter)
- Kuantan (Under Maintenance)
- Kota Kinabalu (Coming Soon)
- Real-time Vehicle Tracking - Live positions of buses and trains
- Stop Search & Information - Find stops by name or location
- Route Discovery - Browse available routes with destinations
- Arrival Predictions - Get real-time arrival times at stops (shape-based, 40-60% more accurate)
- Schedule Information - Get departure times, route schedules, and operating status
- Fare Calculator - Calculate bus fares for BAS.MY and Rapid Penang routes
- Multi-Modal Support - Bus, rail, and ferry services
- Provider Status Monitoring - Check operational status of transit providers
- Location Detection - Automatically detect service areas using geocoding
- 🆕 KTM Komuter Utara - 23 stations from Padang Besar to Ipoh with fare calculation
- 🆕 KTM Intercity - SH and ERT routes (Tumpat - Gemas - JB Sentral)
- 🆕 Penang Ferry - Butterworth to George Town ferry information
- 🆕 API Analytics - View API usage statistics, endpoint metrics, and client tracking
- 🆕 Analytics Dashboard - Visual dashboard with charts for MCP server usage monitoring
- 🆕 Client Identification - MCP identifies itself to middleware for analytics tracking
- 🆕 Auto-Deployment - GitHub Actions workflow for automatic VPS deployment
- 🆕 Multi-Modal Journey Planner (Beta) -
plan_journeycomposes bus + rail + ferry + walking legs across Malaysia (Google Directions engine, locally-computed per-leg fares, AA1 cross-border, free CAT routes) - 🆕 Server-side Places Autocomplete -
places_autocomplete+place_detailsfor resolving Malaysian place names with transit-aware airport overrides - 🆕 Fare Metadata -
get_fare_structuresexposes BAS.MY / Penang / free CAT models so agents can explain how a fare is computed
This MCP server acts as a bridge between AI assistants and the Malaysia Transit Middleware API:
AI Assistant (Claude, GPT, etc.)
↓
Malaysia Transit MCP Server (identifies as "Malaysia-Transit-MCP")
↓
Malaysia Transit Middleware API (tracks usage via X-App-Name header)
↓
Malaysia Open Data Portal (GTFS Static & Realtime)
Client Identification: This MCP automatically sends an X-App-Name: Malaysia-Transit-MCP header with every API request, allowing the middleware to track usage from this MCP separately in its analytics dashboard.
Add this configuration to your MCP client (Claude Desktop, Cursor, Windsurf, etc.):
{
"mcpServers": {
"malaysia-transit": {
"transport": "streamable-http",
"url": "https://mcp.techmavie.digital/malaysiatransit/mcp"
}
}
}npx @modelcontextprotocol/inspector
# Select "Streamable HTTP"
# Enter URL: https://mcp.techmavie.digital/malaysiatransit/mcp# List all available tools
curl -X POST https://mcp.techmavie.digital/malaysiatransit/mcp \
-H "Content-Type: application/json" \
-d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'
# Call the hello tool
curl -X POST https://mcp.techmavie.digital/malaysiatransit/mcp \
-H "Content-Type: application/json" \
-d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"hello","arguments":{}}}'npm installThe MCP server uses environment variables for configuration:
-
MIDDLEWARE_URL(required): Malaysia Transit Middleware API URL- Default:
https://malaysiatransit.techmavie.digital
- Default:
-
INTERNAL_AUTH_TOKEN(required for production): Shared secret for middleware API authentication- Must match
INTERNAL_APP_AUTH_TOKENin the middleware's environment - Generate with:
openssl rand -hex 32
- Must match
-
GOOGLE_MAPS_API_KEY(optional): Google Maps API key for location detection- If not provided, falls back to Nominatim (free but less accurate)
- Get your API key from Google Cloud Console
-
ANALYTICS_RESET_KEY(optional): Admin key forPOST /analytics/reset- If not set, the reset endpoint is disabled (returns 403)
-
ANALYTICS_IMPORT_KEY(optional): Admin key forPOST /analytics/import- If not set, the import endpoint is disabled (returns 403)
To run the MCP server in development mode:
npm run devTo build the MCP server for deployment:
npm run buildList all available transit areas in Malaysia.
Parameters: None
Returns: List of service areas with their IDs, names, and capabilities.
Example:
const areas = await tools.list_service_areas();Get detailed information about a specific area.
Parameters:
areaId(string): Service area ID (e.g., "penang", "klang-valley")
Example:
const info = await tools.get_area_info({ areaId: "penang" });Automatically detect which transit service area a location belongs to using geocoding.
Parameters:
location(string): Location name or place (e.g., "KTM Alor Setar", "Komtar", "KLCC")
Returns: Detected area ID, confidence level, and location details.
Example:
const result = await tools.detect_location_area({ location: "KTM Alor Setar" });
// Returns: { area: "alor-setar", confidence: "high" }Search for stops by name. Use detect_location_area first if unsure about the area.
Parameters:
area(string): Service area IDquery(string): Search query (e.g., "Komtar", "KLCC")
Example:
const stops = await tools.search_stops({
area: "penang",
query: "Komtar"
});Get detailed information about a stop.
Parameters:
area(string): Service area IDstopId(string): Stop ID from search results
Get real-time arrival predictions at a stop.
Parameters:
area(string): Service area IDstopId(string): Stop ID from search results
Returns: Includes a comprehensive disclaimer about prediction methodology, followed by arrival data with:
- Calculation method (shape-based or straight-line)
- Confidence level (high, medium, or low)
- ETA in minutes
- Vehicle information
Prediction Methodology:
- Shape-Based Distance (Preferred): Uses actual route geometry, accurate within ±2-4 minutes
- Straight-Line Distance (Fallback): Conservative estimates with 1.4x multiplier
- Includes GPS speed validation, time-of-day adjustments, and stop dwell time
- Conservative bias: Better to arrive early than miss the bus
Example:
const arrivals = await tools.get_stop_arrivals({
area: "penang",
stopId: "stop_123"
});
// Returns disclaimer + arrival data with confidence levelsFind nearby stops and the routes that serve them.
Parameters:
area(string): Service area IDlocation(string, optional): Place name to geocode (e.g., "KLCC", "Komtar")lat(number, optional): Latitude coordinate (required iflocationnot provided)lon(number, optional): Longitude coordinate (required iflocationnot provided)radius(number, optional): Search radius in meters (default: 500)
Find nearby stops and get real-time arrivals in one call.
Parameters:
area(string): Service area IDlocation(string, optional): Place name to geocodelat(number, optional): Latitude coordinate (required iflocationnot provided)lon(number, optional): Longitude coordinate (required iflocationnot provided)radius(number, optional): Search radius in meters (default: 500)routeFilter(string, optional): Filter arrivals by route number (e.g., "302", "101")
Find nearby stops and get all routes serving those stops in one call.
Parameters:
area(string): Service area IDlocation(string, optional): Place name to geocodelat(number, optional): Latitude coordinate (required iflocationnot provided)lon(number, optional): Longitude coordinate (required iflocationnot provided)radius(number, optional): Search radius in meters (default: 500)
List all routes in an area.
Parameters:
area(string): Service area ID
Get all stops served by a specific route.
Parameters:
area(string): Service area IDrouteId(string): Route ID fromlist_routes
Get detailed route information.
Parameters:
area(string): Service area IDrouteId(string): Route ID from list_routes
Get route path for map visualization.
Parameters:
area(string): Service area IDrouteId(string): Route ID from list_routes
Get real-time vehicle positions.
Parameters:
area(string): Service area IDtype(enum, optional): Filter by type ('bus' or 'rail')
Example:
const vehicles = await tools.get_live_vehicles({ area: "penang" });Check provider operational status.
Parameters:
area(string): Service area ID
Get the next N departures for a specific route (both directions).
Parameters:
area(string): Service area ID (e.g., "ipoh", "seremban", "penang")routeId(string): Route ID from list_routescount(number, optional): Number of departures to return (default: 5)
Example:
const departures = await tools.get_route_departures({
area: "ipoh",
routeId: "A32",
count: 5
});Get the single next departure for a route in a specific direction.
Parameters:
area(string): Service area IDrouteId(string): Route ID from list_routesdirection(enum, optional): 'outbound', 'inbound', or 'loop'
Get all routes serving a specific stop with their next departures.
Parameters:
area(string): Service area IDstopId(string): Stop ID from search_stopscount(number, optional): Number of departures per route (default: 3)
Get the complete daily schedule for a route.
Parameters:
area(string): Service area IDrouteId(string): Route ID from list_routes
Get the origin stop name for a route in a specific direction.
Parameters:
area(string): Service area IDrouteId(string): Route ID from list_routesdirection(enum, optional): 'outbound' or 'inbound'
Check if a route is currently operating based on its schedule.
Parameters:
area(string): Service area IDrouteId(string): Route ID from list_routes
Get all routes available for fare calculation in a specific area.
Parameters:
area(string): Service area ID (e.g., "ipoh", "seremban", "penang", "kangar")
Supported Areas: BAS.MY areas (Ipoh, Seremban, Kangar, Alor Setar, Kota Bharu, Kuala Terengganu, Melaka, Johor, Kuching) and Rapid Penang.
Get all stops on a route with their distances for fare calculation.
Parameters:
area(string): Service area IDrouteId(string): Route ID from get_fare_routes
Calculate the bus fare between two stops on a route.
Parameters:
area(string): Service area IDrouteId(string): Route ID from get_fare_routesfromStop(string): Origin stop IDtoStop(string): Destination stop ID
Returns: Adult, concession, and child fares in MYR with disclaimer.
Example:
const fare = await tools.calculate_fare({
area: "ipoh",
routeId: "A32",
fromStop: "stop_001",
toStop: "stop_015"
});
// Returns: { adult: "1.50", concession: "0.75", child: "FREE" }Calculate the total fare for a multi-leg journey with bus transfers.
Parameters:
area(string): Base service area IDlegs(array): Array of journey legs (max 5), each with:routeId(string): Route ID for this legfromStop(string): Origin stop IDtoStop(string): Destination stop IDareaId(string, optional): Area ID for this leg (for inter-area journeys)
Note: Each bus change requires a separate fare payment (BAS.MY does not have integrated transfers).
Get available directions for a route when calculating fares.
Parameters:
area(string): Service area IDrouteId(string): Route ID from get_fare_routes
Get all 23 KTM Komuter Utara stations (Padang Besar - Butterworth - Ipoh line).
Parameters: None
Returns: Station codes, names, and coordinates.
Example:
const stations = await tools.get_ktm_komuter_stations();
// Returns: [{ code: "PB", name: "Padang Besar", ... }, ...]Calculate KTM Komuter Utara fare between two stations.
Parameters:
from(string): Origin station code (e.g., "BU" for Butterworth)to(string): Destination station code (e.g., "IP" for Ipoh)
Example:
const fare = await tools.calculate_ktm_komuter_fare({
from: "BU",
to: "IP"
});
// Returns: { adult: "12.00", child: "6.00", currency: "MYR" }Get the full KTM Komuter Utara fare matrix showing fares between all station pairs.
Parameters: None
Get departure times for a specific KTM station.
Parameters:
stationName(string): Station name (e.g., "Butterworth", "Ipoh")type(enum): Schedule type -ktm-komuter-utaraorktm-intercity
Example:
const departures = await tools.get_ktm_station_departures({
stationName: "Butterworth",
type: "ktm-komuter-utara"
});Get all KTM stations for a specific schedule type.
Parameters:
type(enum):ktm-komuter-utaraorktm-intercity
Get full KTM schedule data for a specific schedule type.
Parameters:
type(enum):ktm-komuter-utaraorktm-intercity
Find KTM stations near a specific location.
Parameters:
lat(number): Latitude coordinatelon(number): Longitude coordinateradius(number, optional): Search radius in kilometers (default: 10)type(enum):ktm-komuter-utaraorktm-intercity
Get Penang Ferry service overview including route, operator, and general info.
Parameters: None
Get the full Penang Ferry schedule with operating hours and frequency.
Parameters: None
Get the next upcoming Penang Ferry departures from both terminals.
Parameters: None
Get Penang Ferry terminal information including locations and facilities.
Parameters: None
Get Penang Ferry fare information and payment methods.
Parameters: None
Example:
const fare = await tools.get_penang_ferry_fare();
// Returns: { adult: "1.40", child: "0.70", ... }Check the health status of the Malaysia Transit middleware service.
Parameters: None
Get comprehensive debug information about the middleware service.
Parameters: None
Get API usage analytics and statistics from the middleware.
Parameters:
type(enum, optional): Type of analytics to retrieve:summary(default): Overview with requests/hour, error rate, uptimeendpoints: Per-endpoint statisticsareas: Per-service-area statisticscumulative: All-time totalsclients: App/website usage breakdown
Example:
// Get summary analytics
const summary = await tools.get_api_analytics({ type: "summary" });
// Get client usage (see which apps use the API)
const clients = await tools.get_api_analytics({ type: "clients" });Get detailed API usage analytics for a specific service area.
Parameters:
area(string): Service area ID (e.g., "penang", "klang-valley")
Example:
const penangStats = await tools.get_area_analytics({ area: "penang" });Plan a multi-modal public transport journey across Malaysia. Composes bus, rail, ferry, and walking legs via Google Directions in transit mode, then computes per-leg fares locally.
Parameters:
originText(string, optional): Place name (e.g., "Komtar", "Penang Airport", "KL Sentral")originLat/originLng(number, optional): Coordinates (use either text OR lat+lng)destinationText/destinationLat/destinationLng: same shape as originmodeBus/modeRail/modeFerry(boolean, optional): Filter transport modeslanguage(string, optional):en,ms,th
Returns: Array of journey alternatives, each with summary (duration, walking distance, transfer count, fare estimate with hasMissingFare flag), legs[] (mode, route, from/to stops, times, per-leg fare), and walkOnly flag.
Errors: invalid-origin (400), invalid-destination (400), invalid-modes (400), unsupported-inter-area (400), area-out-of-coverage (400), engine-unavailable (503), routing-failed (503). Surface these messages to the user directly; do not retry recursively.
Example:
const plan = await tools.plan_journey({
originText: "Penang Airport",
destinationText: "Komtar",
});Beta caveats: ETAs are scheduled times (use get_stop_arrivals for live ETA at boarding stops). Fares may be missing for some legs (hasMissingFare: true).
List service areas supported by the journey planner, plus engine name and walking/transit config.
Parameters: None
Server-side Google Places autocomplete proxy for Malaysian place names. Returns ranked suggestions including transit-aware overrides (airports snap to canonical bus stops).
Parameters:
input(string, 2-120 chars): Search inputsessiontoken(string, optional): Reuse the same token across autocomplete + details calls for Google Places billinglanguage(string, optional):en,ms,th
Use case: When the user gives an ambiguous place name and you want them to disambiguate before calling plan_journey.
Resolve a placeId from places_autocomplete into coordinates and a formatted address.
Parameters:
placeId(string): Fromplaces_autocomplete(Google place_id oroverride:*synthetic ID)sessiontoken(string, optional): Same session token used in autocompletelanguage(string, optional)
Get the fare structure definitions used by the middleware. Returns BAS.MY distance-based pricing, Rapid Penang staged zones, and free Penang CAT routes (CAT, CT13, C13A/B/C, CT14, CT15).
Parameters: None
Use case: Explain to a user how a fare is computed before calling calculate_fare.
Look up KTM Intercity (SH and ERT routes) fare. NOTE: Currently returns HTTP 501 because Intercity fare data is not yet integrated. The response includes pointers to related endpoints and a link to the official KTMB website. Call this tool to surface the gap honestly to the user rather than guessing a fare.
Parameters: None
Simple test tool to verify server is working.
// 1. Detect area from location
const areaResult = await tools.detect_location_area({
location: "KTM Alor Setar"
});
// 2. Search for your stop
const stops = await tools.search_stops({
area: areaResult.area,
query: "KTM Alor Setar"
});
// 3. Get real-time arrivals
const arrivals = await tools.get_stop_arrivals({
area: areaResult.area,
stopId: stops[0].id
});
// Returns: "Bus K100(I) arrives in 1 minute, Bus K100(O) in 2 minutes"// Get all live vehicles in Penang
const vehicles = await tools.get_live_vehicles({
area: "penang"
});
// Filter by bus only
const buses = await tools.get_live_vehicles({
area: "klang-valley",
type: "bus"
});// List all routes in Klang Valley
const routes = await tools.list_routes({
area: "klang-valley"
});
// Get detailed route information
const routeDetails = await tools.get_route_details({
area: "klang-valley",
routeId: "LRT-KJ"
});This is the PRIMARY use case. Users want to know when their next bus/train will arrive.
Workflow:
1. User asks: "When is the next bus at Komtar?"
2. AI uses: detect_location_area({ location: "Komtar" })
3. AI uses: search_stops({ area: "penang", query: "Komtar" })
4. AI uses: get_stop_arrivals({ area: "penang", stopId: "..." })
5. AI responds: "Bus T101 arrives in 5 minutes, Bus T201 in 12 minutes"
Users want to track their bus in real-time.
Workflow:
1. User asks: "Where is bus T101 right now?"
2. AI uses: detect_location_area({ location: "Penang" })
3. AI uses: get_live_vehicles({ area: "penang" })
4. AI filters for route T101
5. AI responds: "Bus T101 is currently at [location], heading towards Airport"
When a user mentions a location without specifying the area, use location detection:
// User: "When is the next bus at KTM Alor Setar?"
const areaResult = await tools.detect_location_area({
location: "KTM Alor Setar"
});
// Returns: { area: "alor-setar", confidence: "high" }Always search for stops/routes before requesting details:
// ✅ CORRECT
const stops = await tools.search_stops({ area: "penang", query: "Komtar" });
const arrivals = await tools.get_stop_arrivals({
area: "penang",
stopId: stops[0].id
});
// ❌ WRONG - Don't guess stop IDs
const arrivals = await tools.get_stop_arrivals({
area: "penang",
stopId: "random_id"
});Format arrival times in a user-friendly way:
// ✅ GOOD
"Bus T101 arrives in 5 minutes"
"Train LRT-KJ arrives in 2 minutes"
"Next bus: T201 in 12 minutes"
// ❌ BAD
"Arrival time: 2025-01-07T14:30:00Z"
"ETA: 1736258400000"Present multiple arrivals clearly:
"Upcoming arrivals at Komtar:
• T101 → Airport: 5 minutes
• T201 → Bayan Lepas: 12 minutes
• T102 → Gurney: 18 minutes"try {
const arrivals = await tools.get_stop_arrivals({ ... });
} catch (error) {
// Check provider status
const status = await tools.get_provider_status({ area: "penang" });
if (status.providers[0].status !== "active") {
"The transit provider is currently unavailable.
Please try again later or check the official transit app."
}
}- Use location detection when users mention place names
- Always specify area for every tool (except
list_service_areasanddetect_location_area) - Search before details - don't guess IDs
- Handle errors gracefully - providers may have temporary outages
- Format responses clearly - use minutes, not timestamps
- Don't cache real-time data - it updates every 30 seconds
| Area ID | Name | Providers | Transit Types | Fare Calculator |
|---|---|---|---|---|
klang-valley |
Klang Valley | Rapid Rail KL, Rapid Bus KL, MRT Feeder | Bus, Rail | ❌ (Rapid KL fares not modelled) |
penang |
Penang | Rapid Penang (incl. free CAT routes — RM 0), Penang Ferry, KTM Komuter Utara | Bus, Ferry, Rail | ✅ |
kuantan |
Kuantan | Under Maintenance | Bus | ❌ |
ipoh |
Ipoh | BAS.MY Ipoh, KTM Komuter Utara | Bus, Rail | ✅ |
seremban |
Seremban | BAS.MY Seremban, KTM Intercity | Bus, Rail | ✅ |
kangar |
Kangar | BAS.MY Kangar, KTM Komuter Utara | Bus, Rail | ✅ |
alor-setar |
Alor Setar | BAS.MY Alor Setar, KTM Komuter Utara | Bus, Rail | ✅ |
kota-bharu |
Kota Bharu | BAS.MY Kota Bharu, KTM Intercity | Bus, Rail | ✅ |
kuala-terengganu |
Kuala Terengganu | BAS.MY Kuala Terengganu | Bus | ✅ |
melaka |
Melaka | BAS.MY Melaka | Bus | ✅ |
johor |
Johor Bahru | BAS.MY Johor Bahru + Bas Muafakat Johor (BMJ — 41 routes) | Bus | ✅ |
kuching |
Kuching | BAS.MY Kuching | Bus | ✅ |
kota-kinabalu |
Kota Kinabalu | Coming Soon | - | ❌ |
| Service | Route | Stations | Fare Calculator |
|---|---|---|---|
| KTM Komuter Utara | Padang Besar ↔ Butterworth ↔ Ipoh | 23 stations | ✅ |
| KTM Intercity (SH) | JB Sentral ↔ Gemas ↔ Tumpat | Multiple | ❌ (Coming Soon) |
| KTM Intercity (ERT) | JB Sentral ↔ Gemas ↔ Tumpat | Multiple | ❌ (Coming Soon) |
The detect_location_area tool automatically maps common locations to service areas:
| User Says | Area ID |
|---|---|
| Ipoh, Bercham, Tanjung Rambutan, Medan Kidd | ipoh |
| Seremban, Nilai, Port Dickson | seremban |
| George Town, Butterworth, Bayan Lepas, Penang Sentral | penang |
| KLCC, Shah Alam, Putrajaya | klang-valley |
| Kuantan, Pekan, Bandar Indera Mahkota | kuantan |
| Kangar, Arau, Kuala Perlis, Padang Besar | kangar |
| Alor Setar, Sungai Petani, Pendang, Jitra | alor-setar |
| Kota Bharu, Rantau Panjang, Bachok, Machang, Jeli | kota-bharu |
| Kuala Terengganu, Merang, Marang, Setiu | kuala-terengganu |
| Melaka, Tampin, Jasin, Masjid Tanah | melaka |
| Johor Bahru, Iskandar Puteri, Pasir Gudang, Kulai | johor |
| Kuching, Bau, Serian, Bako, Siniawan, Matang | kuching |
The MCP server is deployed at:
- Endpoint:
https://mcp.techmavie.digital/malaysiatransit/mcp - Health Check:
https://mcp.techmavie.digital/malaysiatransit/health - Analytics Dashboard:
https://mcp.techmavie.digital/malaysiatransit/analytics/dashboard - Analytics API:
https://mcp.techmavie.digital/malaysiatransit/analytics - Transport: Streamable HTTP
The MCP server includes a built-in analytics dashboard that tracks:
- Total requests and tool calls
- Tool usage distribution (doughnut chart)
- Hourly request trends (last 24 hours)
- Requests by endpoint (bar chart)
- Top clients by user agent
- Recent tool calls feed
The dashboard auto-refreshes every 30 seconds.
To deploy your own instance, see the deployment guide.
# Using Docker
docker compose up -d --build
# Or run directly
npm run build
npm run start:httpThis repository includes a GitHub Actions workflow for automatic VPS deployment. When you push to main, the server automatically redeploys.
To set up auto-deployment, add these secrets to your GitHub repository:
VPS_HOST- Your VPS IP addressVPS_USERNAME- SSH username (e.g.,root)VPS_PORT- SSH port (e.g.,22)VPS_SSH_KEY- Private SSH key for authenticationGOOGLE_MAPS_API_KEY- Google Maps API key for geocodingINTERNAL_AUTH_TOKEN- Shared auth token (must match middleware'sINTERNAL_APP_AUTH_TOKEN)ANALYTICS_RESET_KEY(optional) - Enables/analytics/resetadmin endpointANALYTICS_IMPORT_KEY(optional) - Enables/analytics/importadmin endpoint
If you can't connect to the middleware:
- Verify your
MIDDLEWARE_URLis correct - Ensure the middleware is running and accessible
- Check network connectivity
- Verify
INTERNAL_AUTH_TOKENmatches middlewareINTERNAL_APP_AUTH_TOKEN(if middleware auth is enabled) - Test middleware directly:
curl https://your-middleware-url/api/areas
If tools return empty data:
- Check if the service area is operational using
get_provider_status - Verify the area ID is correct using
list_service_areas - Check middleware logs for errors
Real-time data depends on the upstream GTFS providers:
- Use
get_provider_statusto check provider health - Some providers may have temporary outages
- Check the middleware logs for API issues
If location detection returns incorrect results:
- Ensure
GOOGLE_MAPS_API_KEYis set in environment variables - Check Google Cloud Console for API quota limits
- Verify the API key has Geocoding API enabled
- Falls back to Nominatim if Google Maps fails
- Node.js: >= 18.0.0
- Malaysia Transit Middleware: Running instance (local or deployed)
- Google Maps API Key: Optional, for enhanced location detection
malaysiatransit-mcp/
├── src/
│ ├── index.ts # Main MCP server entry point
│ ├── http-server.ts # Streamable HTTP server with analytics
│ ├── transit.tools.ts # Transit tool implementations
│ ├── geocoding.utils.ts # Location detection utilities
│ ├── inspector.ts # MCP Inspector entry point
│ └── server.ts # HTTP server for testing
├── deploy/
│ ├── DEPLOYMENT.md # VPS deployment guide
│ └── nginx-mcp.conf # Nginx reverse proxy config
├── .github/
│ └── workflows/
│ └── deploy-vps.yml # GitHub Actions auto-deploy
├── docker-compose.yml # Docker deployment config
├── Dockerfile # Container build config
├── package.json # Project dependencies
├── tsconfig.json # TypeScript configuration
├── .env.sample # Environment variables template
├── README.md # This file
└── LICENSE # MIT License
- Malaysia Open Data MCP - MCP for Malaysia's open data portal
Contributions are welcome! Please feel free to submit pull requests or open issues.
MIT - See LICENSE file for details.
- Malaysia Open Data Portal for GTFS data
- Prasarana Malaysia for Rapid KL services
- BAS.MY for regional bus services
- Model Context Protocol for the MCP framework
- Google Maps Platform for geocoding services
- OpenStreetMap Nominatim for fallback geocoding
Made with ❤️ by Aliff