Fleet management with AI: your agent reads the fleet.
The VehDB MCP server turns your curated fleet list into a native tool for Claude, Cursor and any MCP client. The agent reads the list your team maintains — then researches every vehicle against the full catalog, in the same conversation.
OAuth2 for AI providers, Bearer tokens for everything else. Free tier included.
Which vans do we own, and what tires do they take?
You own 2 vans. OEM fitments:
- · 2022 Transit 250 → LT245/75R16
- · 2020 Sprinter 2500 → LT245/75R16
via my_lists + tire_sizes_for_vehicle
How it works
Three steps, no integration project
1 · Curate the fleet
Your team saves vehicles into lists in the dashboard — statuses (owned, wishlist, considering) and notes. This stays human: the list is your source of truth.
2 · Connect the MCP server
In Claude, add VehDB from the connector directory and sign in. Other clients take one URL: OAuth2 sign-in, or a plain Bearer token. Two minutes, no SDK.
3 · Ask in plain language
The agent calls my_lists to read the fleet, then search_vehicles, get_vehicle and tire_sizes_for_vehicle to research each row — and answers in one message.
In practice
Ask things a spreadsheet can't answer
Every prompt below chains your list with the catalog — the agent does the lookups, you keep the answer.
“What’s on our fleet right now, grouped by status?”
my_lists
“Which wishlist trucks are AWD diesels under $0.09/mile in fuel?”
my_lists · search_vehicles
“Pull OEM tire sizes for every vehicle we own.”
my_lists · tire_sizes_for_vehicle
“Compare the EPA fuel cost of our three “considering” picks.”
my_lists · get_vehicle
Beyond your list
Every vehicle question, answered from the records
The same agent answers about any vehicle, not only the ones you saved. Live examples:
recalls_for_vehicle
- Recall campaigns
- 23
- Open investigations
- 5
- Newest
- 2026-05-04
From today's VehDB data: NHTSA, EPA and the vehicle catalogue.
model_year_scorecard
- Best year
- 2022
- Worst year
- 2007
- Years compared
- 25
From today's VehDB data: NHTSA, EPA and the vehicle catalogue.
search_vehicles
get_vehicle
- Combined MPG
- 51
- Fuel cost / yr
- $1,250
- CO₂
- 173 g/mi
From today's VehDB data: NHTSA, EPA and the vehicle catalogue.
tire_sizes_for_vehicle
- Most common
- 275/60R20
- OEM sizes
- 8
From today's VehDB data: NHTSA, EPA and the vehicle catalogue.
vehicle_insights
- Fatal-crash vehicles
- 14
- Occupant deaths
- 8
From today's VehDB data: NHTSA, EPA and the vehicle catalogue.
vehicle_insights
- Ford
- 135 campaigns
- Vehicles in them
- 12,632,844
From today's VehDB data: NHTSA, EPA and the vehicle catalogue.
my_lists
recalls_for_vehicle
decode_vin
tire_sizes_for_vehicle
recalls_for_vehicle
Enriched data
What comes back with every answer
One vehicle, every public record joined to it: specs from the catalogue, economy and emissions from the EPA, safety from NHTSA. Counts are live.
- engine
- drive
- transmission
- body
- doors
- markets
search_vehicles
- mpg city/highway
- EV range
- annual fuel cost
- CO₂ g/mi
get_vehicle
- smog rating
- SmartWay
- greenhouse-gas score
get_vehicle
- campaign
- component
- remedy
- do-not-drive
- build window
recalls_for_vehicle
- by component
- crashes
- fires
- injuries
- percentile
complaints_for_vehicle
- service bulletins
- warranty extensions
- over-the-air updates
- components
service_bulletins_for_vehicle
- open investigations
- NCAP stars
- front/side/rollover
recalls_for_vehicle
- fatal crashes
- occupant deaths
- injury-crash estimates
vehicle_insights
- OEM size
- alternate sizes
- diameter delta
- load & speed index
tire_sizes_for_vehicle
- make, model, year
- engine
- plant
- matching VehDB trims
decode_vin
- HIN decode
- builder
- model year
- USCG recalls
decode_hin
- plant
- build week
- tire age
- covering recalls
tire_dot_check
- gasoline grades
- diesel
- electricity by state
- annual fuel cost
energy_prices
- best & worst year
- complaint percentile
- flags
model_year_scorecard
No magic
What the agent can and can't do
Trust comes from boundaries. Here's exactly where they are.
| Capability | How | Status |
|---|---|---|
| Read the fleet list | my_lists (MCP) | Live |
| Research each vehicle — specs, EPA fuel economy | search_vehicles · get_vehicle | Live |
| OEM & alternate tire fitments per vehicle | tire_sizes_for_vehicle | Live |
| Recall exposure across the fleet | GET /v1/lists/recalls + daily email digest | Via API |
| Add or remove fleet vehicles | dashboard | Human, by design |
The agent never invents your fleet — it can only read the list your team actually curated.
Setup
One config block away
Any MCP client: paste the server, drop in a token from the dashboard, ask your first fleet question.
{ "mcpServers": { "vehdb": {
"url": "https://mcp.vehdb.com/mcp",
"headers": { "Authorization": "Bearer TOKEN" } } } }
Tools your agent gets: search_vehicles · get_vehicle · my_lists · tire_sizes_for_vehicle
Give your AI a fleet to work with.
Create the list in the dashboard, connect the MCP server, ask your first question. Free tier included.
100 free API calls every month · REST API + MCP server on every plan · No card to start · See pricing