Model Context Protocol moderate risk

Geoserver

GeoServer MCP server connects LLMs to GeoServer's REST API, enabling AI-driven geospatial data interaction and management, including querying, visualization, and manipulation.

Connections & Capabilities

Connects To

GitHubDocker

Capabilities

readwriteexecadmin

Quickstart

Install

npx -y @smithery/cli install @mahdin75/geoserver-mcp --client claude

Config

{
  "mcpServers": {
    "geoserver-mcp": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "GEOSERVER_URL=http://localhost:8080/geoserver",
        "-e",
        "GEOSERVER_USER=admin",
        "-e",
        "GEOSERVER_PASSWORD=geoserver",
        "-p",
        "8080:8080",
        "mahdin75/geoserver-mcp"
      ]
    }
  }
}

Exposed MCP Tools (6)

safe

list_workspaces

Lists available workspaces in the GeoServer instance.

Read-only operation; no data modification.

moderate

create_workspace

Creates a new workspace in GeoServer.

Creates a new workspace, but doesn't modify existing data.

moderate

create_datastore

Creates a new datastore in a specified workspace.

Adds a new datastore, but doesn't modify existing data.

moderate

create_featurestore

Creates a new featurestore in the specified workspace.

Adds a new featurestore, but doesn't modify existing data.

moderate

create_gpkg_datastore

Creates a GeoPackage datastore in GeoServer.

Creates a new datastore; no modification of existing data.

moderate

create_shp_datastore

Creates a shapefile datastore in GeoServer.

Creates a new datastore; no modification of existing data.

Safety Assessment

The GeoServer MCP server offers a convenient way to integrate LLMs with geospatial data. However, the ability to perform write operations and the reliance on GeoServer's authentication mechanisms necessitate careful configuration and monitoring. It's relatively safe for read-only operations, but risky if write access is granted without proper controls.

Access to GeoServer is controlled by username/password.
MCP limits the scope of LLM interaction to defined tools.
Can be deployed in a sandboxed environment using Docker.
Clear separation of concerns between LLM and GeoServer instance.

Improperly configured GeoServer instances can expose sensitive data.
Write operations can modify geospatial data.
No built-in rate limiting or usage controls.
Vulnerable to injection attacks if input sanitization is insufficient.
Default credentials are provided in examples, which is a security risk.