MCP SERVERAdminister a GeoServer instance from an assistant — workspaces, datastores, layers, SLD styles, layer groups, users and server health — over its REST API.
LABSCO SUMMARY
Its reach is the story here: standing up a layer normally means clicking through store creation, feature-type publication and style assignment across several admin pages, and this collapses that into a sequence an assistant can run. The style generators are the neat part — create_classified_featurestyle turns a column and a color ramp into a working SLD, so you get a thematic map without hand-writing XML. It is still marked beta, so expect the surface to keep moving.
A server connecting an LLM to the GeoServer REST API: 56 tools covering the full admin surface — workspaces and stores, layers and feature types, SLD style generation, layer groups, security, and server status — plus WMS map rendering and CQL feature queries.
A running GeoServer and admin credentials: GEOSERVER_URL (e.g. http://localhost:8080/geoserver), GEOSERVER_USER and GEOSERVER_PASSWORD. Python, installed from the geoserver-mcp PyPI package with the uv package manager, or run via Docker. A --storage path can root the file paths that create_gpkg_datastore and create_shp_datastore resolve against.
Follow the published steps — uv pip install geoserver-mcp
USE CASES
Create a workspace, register a shapefile datastore, publish its feature type, and confirm the layer is there — the whole path from a file on disk to a served layer.
Generate a classified or categorized style from a column and a color ramp, assign it to the layer, and render a WMS image over a bounding box to check how it looks.
Run a CQL filter against a layer, asking for just the properties you need and capping the feature count, without exporting the data first.
Create a group, add the layers it should contain with their styles, and update its title and metadata before it goes live.
Check server and system status, read the version, and reload or reset the catalog after a configuration change lands.
Counted once with only startup setup. Other versions, modes or optional credentials may expose a different list.
add_layer_to_layergroupAdd a specific layer to a layer group.
Args: layer_name (str): Layer to add. layer_workspace (str): Workspace for the layer. layergroup_name (str): Target group name. layergroup_workspace (str, optional): Workspace for the group. Returns: dict: Status and any info.
create_catagorized_featurestyleCreate a categorized style for features (polygon, line, point) using column values.
Args: style_name (str): New style name. column_name (str): Column to base categories on. column_distinct_values: Values (categories) to style. workspace (str, optional): Workspace for the style. color_ramp (str, optional): Color ramp. geom_type (str, optional): Geometry type. Returns: dict: API response.
create_classified_featurestyleCreate a classified style for features using distinct column values/classes.
Args: style_name (str): New style name. column_name (str): Which column. column_distinct_values: Classes/categories for styling. workspace (str, optional): Workspace. color_ramp (str, optional): Ramp. geom_type (str, optional): Geometry type. Returns: dict: API response.
create_coveragestoreCreate a new coveragestore in a workspace.
Args: workspace (str): Workspace name. name (str): Coveragestore name. params (dict): Parameters (url, type, etc.)
Returns: dict: Creation response.
create_coveragestyleCreate a raster coverage style (colormap, ...)
Args: style_name (str): Name for the style. params (dict): Colormap/settings for style. Returns: dict: Server response.
create_datastoreCreate a new datastore in the given workspace.
Args: workspace (str): Name of the workspace to add the datastore to. name (str): The datastore name. params (dict): Connection parameters (host, dbtype, etc.).
Returns: dict: Details of the created datastore or API response.
create_featurestoreCreate a new featurestore in the given workspace.
Args: workspace (str): Target workspace. name (str): Featurestore name. params (dict): Connection/configuration parameters.
Returns: dict: Details of the created featurestore.
create_gpkg_datastoreCreate a GeoPackage (GPKG) datastore.
Args: workspace (str): Workspace name. name (str): Datastore name. file_path (str): Path to the .gpkg file (relative to --storage if provided) Returns: dict: Creation result.
create_layerCreate a new layer in GeoServer.
Args: workspace: The workspace for the new layer layer: The name of the layer to create data_store: The data store to use source: The source data (file, table name, etc.)
Returns: Dict with status and layer information
create_layergroupCreate a new layer group with specific layers and (optionally) styles.
Args: workspace (str): The workspace for the group. name (str): Name of the layer group. layers (list): List of layers to include. styles (list, optional): List of styles for layers. Returns: dict: Creation result.
create_outline_featurestyleCreate a simple outline-only style for features.
Args: style_name (str): Style name. outline_color (str): HEX outline color. workspace (str, optional): Workspace. Returns: dict: API creation response.
create_shp_datastoreCreate an ESRI Shapefile datastore.
Args: workspace (str): Workspace in GeoServer. name (str): New datastore name. file_path (str): Path to .shp or zipped shapefile (relative to --storage if provided) Returns: dict: Creation result.
create_styleCreate a new SLD style in GeoServer.
Args: name: Name for the style sld: SLD XML content workspace: Optional workspace for the style
Returns: Dict with status and style information
create_userCreate a new user for GeoServer security.
Args: username (str): New user's name. password (str): Password for the user. enabled (bool): Whether account is enabled on creation. properties (dict, optional): Additional user properties.
Returns: dict: API response on user creation.
create_usergroupCreate a new user group.
Args: name (str): Name for the user group. users (list, optional): Users to add initially. Returns: dict: Creation outcome.
create_workspaceCreate a new workspace in GeoServer.
Args: workspace: Name of the workspace to create
Returns: Dict with status and result information
delete_coveragestoreDelete a coveragestore from a workspace.
Args: workspace (str): Workspace name. name (str): Coveragestore name. Returns: dict: Response from API about deletion result.
delete_layergroupDelete a layer group from a workspace.
Args: workspace (str): Workspace. name (str): Group to delete. Returns: dict: Result of the delete operation.
delete_resourceDelete a resource from GeoServer.
Args: resource_type: Type of resource to delete (workspace, layer, style, etc.) workspace: The workspace containing the resource name: The name of the resource
Returns: Dict with status and result information
delete_userDelete a user by name.
Args: username (str): User to remove. Returns: dict: Deletion result.
delete_usergroupDelete a user group.
Args: name (str): Name of group to remove. Returns: dict: Result of group deletion.
edit_featuretypeEdit the settings of a feature type in a store.
Args: workspace (str): Workspace containing store. store_name (str): Store name. featuretype (str): Feature type to modify. **kwargs: Any updatable featuretype attributes. Returns: dict: API update result.
generate_mapGenerate a map image using WMS GetMap.
Args: layers: List of layers to include (format: workspace:layer) styles: Optional styles to apply (one per layer) bbox: Bounding box [minx, miny, maxx, maxy] width: Image width in pixels height: Image height in pixels format: Image format (png, jpeg, etc.)
Returns: Dict with map information and URL
get_all_usergroupsReturn all user groups.
Returns: list: Names or info dicts for all groups.
get_all_usersList all users in the GeoServer instance.
Returns: list: Usernames or user info dicts.
get_coveragestoreGet details about a single coveragestore.
Args: workspace (str): Workspace. name (str): Coveragestore. Returns: dict: Metadata for the coveragestore.
get_coveragestoresGet all coveragestores in a workspace.
Args: workspace (str): Name of the workspace. Returns: list: List of coveragestores.
get_datastoreGet a specific datastore by name.
Args: workspace (str): Workspace name. name (str): Datastore name. Returns: dict: Datastore description.
get_datastoresList all datastores in the given workspace.
Args: workspace (str): Workspace. Returns: list: List of datastores for the workspace.
get_feature_attributeGet feature attribute schema/details.
Args: workspace (str): Workspace. store_name (str): Store containing layer. featuretype (str): The layer or featuretype name. Returns: dict: Attribute info.
get_featuretypesList all feature types in a given store.
Args: workspace (str): Workspace. store_name (str): Store name. Returns: list: Featuretype metadata/list.
get_layer_infoGet detailed information about a layer.
Args: workspace: The workspace containing the layer layer: The name of the layer
Returns: Dict with layer metadata
get_layergroupGet a layer group from a workspace.
Args: workspace (str): Workspace to search. name (str): Name of the group. Returns: dict: Layer group metadata/details.
get_layergroupsList all layer groups in a workspace.
Args: workspace (str): Workspace. Returns: list: Layer group names or info dicts.
get_manifestGet GeoServer manifest metadata/details.
Returns: dict: Manifest metadata from GeoServer.
get_statusObtain general server status.
Returns: dict: Status/health report.
get_system_statusGet system status overview/info from GeoServer.
Returns: dict: System info (CPU, memory, etc).
get_versionFetch GeoServer version string.
Returns: str: GeoServer version.
list_layersList layers in GeoServer, optionally filtered by workspace.
Args: workspace: Optional workspace to filter layers
Returns: List of layer information dictionaries
list_workspacesList available workspaces in GeoServer.
modify_userModify an existing user's properties (password, enabled state, properties, etc).
Args: username (str): Name to update. **kwargs: Attributes to set/update. Returns: dict: Response message.
publish_featurestorePublish an existing featurestore.
Args: workspace (str): Target workspace. store_name (str): Featurestore name. params (dict): Publication settings (table, type, etc). Returns: dict: API result.
publish_featurestore_sqlviewPublish a featurestore using a SQL view definition.
Args: workspace (str): Workspace context. store_name (str): Featurestore. params (dict): Featurestore details. sqlview_params (list): SQL view parameter dicts. Returns: dict: Publication outcome.
publish_styleAssign/publish a style to a layer.
Args: layer_name (str): The target layer. style_name (str): The style to apply. workspace (str): Workspace context. Returns: dict: API response from publish.
publish_time_dimension_to_coveragestoreAdd or update a time dimension for a coverage store (for time series).
Args: store_name (str, optional): Target coverage store. workspace (str, optional): Workspace. presentation (str, optional): Presentation style. units (str, optional): Time units. default_value (str, optional): Default value for the time dimension. content_type (str, optional): Payload content type.
Returns: dict: Response from GeoServer.
query_featuresQuery features from a vector layer using CQL filter.
Args: workspace: The workspace containing the layer layer: The layer to query filter: Optional CQL filter expression properties: Optional list of properties to return max_features: Maximum number of features to return
Returns: GeoJSON FeatureCollection with query results
reload_geoserverReload catalog and config from disk.
Returns: str: Result message from reload.
remove_layer_from_layergroupRemove a layer from a group.
Args: layer_name (str): Layer. layer_workspace (str): Layer workspace. layergroup_name (str): Group. layergroup_workspace (str, optional): Group workspace. Returns: dict: Deletion result.
reset_geoserverReset all GeoServer caches/connections.
Returns: str: Result message from reset operation.
style_catagorize_xmlGenerate SLD for categorized vector style (SVG fill/block display). Args: column_name (str): Column for category. values (list): Category values. color_ramp (str, optional): Color ramp. geom_type (str, optional): Geometry. Returns: str: SLD XML.
style_classified_xmlGet SLD XML for classified vector style. Args: style_name (str): Name of style. column_name (str): Column to use for class. values (list): Scalar classes. color_ramp (str, optional): Colors. geom_type (str, optional): Geometry type. Returns: str: SLD XML.
style_coverage_style_colormapentryGenerate color map entries for raster SLD. Args: color_ramp: List/dict of colors. min_value (float): Minimum value. max_value (float): Maximum value. number_of_classes (int, optional): Number of classes. Returns: List or dict: Colormap structure for SLD.
style_coverage_style_xmlGenerate XML for raster/coverage SLD. Args: color_ramp: Color ramp (list, dict...) style_name: Name of style. cmap_type: Colormap type. min_value: Minimum. max_value: Maximum. number_of_classes: Number of classes. opacity: Opacity. Returns: str: XML for the style.
style_outline_only_xmlXML for outline-only style for a geometry. Args: color (str): Outline color (hex). width (float): Line width. geom_type (str): Geometry type ('polygon', etc). Returns: str: SLD XML for outline style.
update_layergroupUpdate a layer group's details and configuration.
Args: layergroup_name (str): The group to update. title (str, optional): New title. abstract_text (str, optional): Abstract/description. formats (str, optional): Response format (default html). metadata (list, optional): Extra metadata entries. keywords (list, optional): Associated keywords. Returns: dict: Result message.
update_serviceUpdate selected OGC service options.
Args: service (str): Service (e.g., 'wfs', 'wms'). options (dict): Options/params to change. Returns: str: Result message from service update.
BEFORE YOU INSTALL
56 tools when we connected. generate_map ships but is missing from the README's tool tables, and the user and group tools ship although the README still lists security as planned. Two names must be spelled as shipped: create_catagorized_featurestyle and style_catagorize_xml.
All 4 documented setups run the server on your machine.
CONTEXT COST
Rated by the size of the 56 tool definitions that ride along on every turn, not by how many there are
INSTALLATION
Run these in the folder you want the server in (for example a new geoserver-mcp folder); uv venv creates the .venv there:
pip install uv
uv venv --python=3.10
uv pip install geoserver-mcp{
"mcpServers": {
"geoserver-mcp": {
"command": "<YOUR_PATH_TO_GEOSERVER_MCP>/.venv/bin/geoserver-mcp",
"args": [
"--url",
"http://localhost:8080/geoserver",
"--user",
"admin",
"--password",
"geoserver"
]
}
}
}<YOUR_PATH_TO_GEOSERVER_MCP> is the full path of the folder where you ran uv venv (on Windows the command is C:\path\to\geoserver-mcp\.venv\Scripts\geoserver-mcp). Set --url, --user and --password to your GeoServer's address and admin login.
{
"mcpServers": {
"geoserver-mcp": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e",
"GEOSERVER_URL=http://localhost:8080/geoserver",
"-e",
"GEOSERVER_USER=admin",
"-e",
"GEOSERVER_PASSWORD=geoserver",
"mahdin75/geoserver-mcp"
]
}
}
}Set GEOSERVER_URL, GEOSERVER_USER and GEOSERVER_PASSWORD to your GeoServer's address and admin login. The URL is resolved inside the container, where localhost is the container itself, so use an address of your GeoServer that the container can reach.
MCP servers
jagan-shanmugam
AiAgentKarl
jjsantos01
birdingkit