About Pumperly Mcp
Pumperly Mcp bridges a Pumperly instance to MCP clients, providing read-only access to fuel and EV charging information in about 40 countries. It helps agents find stations near a place or along a route, with results summarized by price and distance.
Key features
- 01Six read-only search and route tools
- 02Find fuel or charging along a drive
- 03Search stations by location, distance, or price
- 04Four resources for fuel, config, stats, and rates
- 05stdio and authenticated HTTP transports
Use cases
- 01Find the cheapest fuel between two places
- 02Locate nearby charging stations by price
- 03List stations inside a geographic area
- 04Get driving distance and time with stops
MCP support
The server exposes six read-only tools and four resources to MCP clients. It supports stdio and streamable HTTP transports, and can connect to the public Pumperly instance or one configured with PUMPERLY_URL.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
npx -y pumperly-mcpPUMPERLY_URLPumperly instance URL (default https://pumperly.com); set it to use your own instanceLocal servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.
Tools
Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.
Details
From the project README
pumperly-mcp
A small bridge that exposes any Pumperly instance as an MCP server, so an AI agent can find the cheapest fuel or charging near a place or along a drive, in about 40 countries, mostly in Europe.
Features
-
Six read-only tools:
Tool What it answers cheapest_fuel_on_routeWhere to fill up between two places, in one call find_nearest_stationsStations near a place name or coordinates, by distance or price get_stations_in_areaStations inside a bounding box, cheapest first calculate_routeDriving distance and time between places, with up to 5 stops find_route_stationsStations along a route geometry you already have geocodeCoordinates for a place name -
Four resources:
pumperly://fuel-types,pumperly://config,pumperly://stats,pumperly://exchange-rates. -
Short answers: station lists are sorted, cut to a limit and summarised (min, median and max price), so a search costs about 1 KB of context.
-
MCP 2026-07-28 and every earlier version back to 2024-11-05, on the same binary: new clients use
server/discover, older onesinitialize. -
Two transports: stdio, and streamable HTTP on
/mcpwith a bearer token and origin checks. -
Works with the public instance at pumperly.com or your own: set
PUMPERLY_URL. No API key.
Every tool only reads, so there is no read-only switch to set.
Install
The npm package downloads the release binary for your platform, checks its SHA-256 against the checksums shipped in the package, and runs it over stdio. It needs Node.js 18 or later.
Claude Code:
claude mcp add pumperly -- npx -y pumperly-mcp
Claude Desktop (claude_desktop_config.json) and Cursor (~/.cursor/mcp.json):
{
"mcpServers": {
"pumperly": { "command": "npx", "args": ["-y", "pumperly-mcp"] }
}
}
VS Code (.vscode/mcp.json):
{
"servers": {
"pumperly": { "type": "stdio", "command": "npx", "args": ["-y", "pumperly-mcp"] }
}
}
Codex (~/.codex/config.toml):
[mcp_servers.pumperly]
command = "npx"
args = ["-y", "pumperly-mcp"]
Docker, over stdio:
docker run -i --rm -e TRANSPORT=stdio drumsergio/pumperly-mcp:v0.2.0
To use your own Pumperly instance, add PUMPERLY_URL to the environment ("env": { "PUMPERLY_URL": "https://..." } in the JSON configs). For the HTTP transport, see Getting started.
Documentation
- Getting started: npm, Docker Compose, building from source, connecting a client
- Configuration: environment variables
- Usage: tools, resources, the request flow, testing with the Inspector
- Related projects: the Pumperly family, other MCP servers, registry listings
License
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.