connected-car-mcp
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@connected-car-mcp Which vehicles in the fleet need attention right now, and why?"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
connected-car-mcp
An MCP (Model Context Protocol) server over a synthetic connected-vehicle fleet: telemetry, rule-based anomaly detection, and maintenance recommendations, exposed as five narrow tools instead of one open-ended query interface.
Built as a small, self-contained illustration of a specific design habit: deciding what belongs behind a tool boundary, and logging every call across it. Runs entirely on synthetic data generated locally — no external API, no account, no proprietary source.
Why it's shaped this way
The whole dataset could be exposed through a single run_query(sql: str)
tool. That's the wrong shape for an agent to call reliably: it pushes
schema-learning onto the model at call time, and there's no way to scope or
audit "what can be asked" per capability. Instead:
Tool | Contract |
| Enumerate the fleet |
| Raw readings for one vehicle, time-bounded |
| Latest snapshot + fleet averages |
| Rule-based flags: overheating, low battery, fault codes, harsh driving |
| Prioritized actions for one vehicle |
A model composes these — summary → pick a flagged vehicle → pull its
telemetry → get a recommendation — rather than writing free-form queries
against raw rows. It also makes the audit story trivial: there are only
five well-defined calls to log, so audit_log.jsonl (see
connected_car_mcp/audit.py) is one line per
call — timestamp, tool, arguments, duration, success/failure — with no
custom logic per tool. A production deployment would emit the same record
as structured logs via MCPServer's middleware hook (which sees every
JSON-RPC call, tool or resource) rather than a local file; the decorator
here keeps the demo runnable with zero extra infrastructure.
Anomaly thresholds are simple and explainable (engine_temp_c >= 110, not
a trained model) on purpose — a fleet monitor's flags need to be
auditable by a human, not just accurate.
Related MCP server: incident-mcp
Data
data/generate_telemetry.py generates a deterministic (fixed-seed),
fully synthetic dataset: 12 vehicles, readings every 10 minutes over 3 days.
A few vehicles are seeded with faults so the anomaly detector has real
signal to find:
CCV-004 — engine temperature ramps into critical range (cooling system failure)
CCV-009 — battery voltage degrades over time (failing battery/alternator)
CCV-002, CCV-011 — intermittent DTC fault codes
CCV-006 — occasional harsh-driving speed spikes
data/telemetry.csv is committed so the repo runs immediately; regenerate
it with:
python data/generate_telemetry.pyRunning it
python -m venv .venv
.venv/Scripts/activate # .venv/bin/activate on macOS/Linux
pip install -r requirements.txt
python -m connected_car_mcp.server # starts the MCP server over stdioTo try it from Claude Desktop or another MCP client, point it at the
module with cwd set to the repo root, e.g. in claude_desktop_config.json:
{
"mcpServers": {
"connected-car-fleet": {
"command": "python",
"args": ["-m", "connected_car_mcp.server"],
"cwd": "/path/to/connected-car-mcp"
}
}
}Then ask something like "Which vehicles in the fleet need attention right
now, and why?" — the model will call fleet_health_summary, follow up
with detect_anomalies on the flagged vehicles, and can call
get_maintenance_recommendations to turn that into next actions.
Tests
pip install pytest
pytest tests/Covers the data layer directly (fleet size, unknown-vehicle handling, and that the seeded faults actually get flagged) rather than round-tripping through the MCP protocol layer.
Project layout
connected_car_mcp/
server.py MCP tool + resource definitions
data_store.py Query layer over the telemetry CSV (pandas)
audit.py Per-call audit log decorator
data/
generate_telemetry.py Synthetic dataset generator
telemetry.csv Generated dataset (committed)
tests/
test_data_store.pyLicense
MIT — see LICENSE.
Available Tools
5 toolsdetect_anomaliesA
List rule-based anomalies (overheating, low battery, fault codes, harsh driving) across the fleet, or for a single vehicle_id if given.
Inputs: vehicle_id. Returns: result
fleet_health_summaryA
Get a fleet-wide snapshot: latest reading per vehicle, fleet averages, and which vehicles currently have an active anomaly flag.
Inputs: no named parameters.
get_maintenance_recommendationsB
Get prioritized maintenance recommendations for one vehicle, combining service-interval mileage with any active anomaly flags.
Inputs: vehicle_id (required). Returns: result
get_vehicle_telemetryA
Get raw telemetry readings for one vehicle, optionally bounded by an
ISO-8601 start/end timestamp. Returns at most limit readings, most
recent first within that window.
Inputs: vehicle_id (required), end, limit, start. Returns: result
list_vehiclesA
List every vehicle in the fleet with its first/last telemetry timestamp.
Inputs: no named parameters. Returns: result
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 5 tools
Every tool targets a distinct concern: listing vehicles, retrieving raw telemetry, fleet-wide health snapshot, anomaly detection, and maintenance recommendations. The overlap between fleet_health_summary and detect_anomalies is minimal because one is a snapshot and the other is a detailed listing.
Most tools follow a clear verb_noun pattern (list_vehicles, get_vehicle_telemetry, detect_anomalies, get_maintenance_recommendations). The exception is fleet_health_summary, which lacks a verb prefix, creating a minor inconsistency in the naming scheme.
Five tools is well-scoped for a connected-car MCP, covering the core operations of vehicle listing, telemetry retrieval, fleet health, anomaly detection, and maintenance recommendations without redundancy or bloat.
The set covers the primary read/analysis workflows: listing vehicles, fetching raw data, summarizing health, detecting anomalies, and recommending maintenance. A minor gap is the lack of a tool for direct vehicle metadata (e.g., model, year), but this is not critical for the apparent monitoring/analytics purpose.
Maintenance
Related MCP Connectors
Public MCP digital twin with synthetic systems and an agent firewall. No customer data.
Governed data discovery, exact queries, decisions, simulations, and runtime utilities over MCP.
A paid remote MCP for AI SDK eval dashboard, built to return verdicts, receipts, usage logs, and aud
One OAuth connector for your team's remote MCP servers, with Tool Packs, guardrails and logs.
Related MCP Servers
- AlicenseAqualityCmaintenanceExposes live industrial IoT telemetry to any MCP client, streaming simulated sensor data from a fleet of machines and detecting anomalies, with the ability to inject faults on demand.4MIT
- AlicenseNot gradedqualityCmaintenanceEnables incident detection and analysis by identifying anomalies in metric time series and surfacing root-cause candidates and recommended actions. Supports both mock (synthetic) and VictoriaMetrics backends with identical MCP tool contracts for seamless development-to-production switching.MIT
- AlicenseAqualityDmaintenanceExposes digital twin device data capabilities as MCP tools, allowing MCP clients to query device status, read real-time metrics, fetch time-series data, and check alerts. Includes a simulated PLC driver with a clean interface for connecting real devices via OPC-UA, Modbus, or gateway APIs.6MIT
- AlicenseNot gradedqualityDmaintenanceEnables Claude Code to query fleet telemetry data for device health summaries, anomaly detection, capacity breach forecasts, and incident history via narrow MCP tools.MIT