User MCP server

class datarobot._experimental.models.user_mcp_server_deployment.EnumAPIRepresentationConverter

class datarobot._experimental.models.user_mcp_server_deployment.PromptInUserMCPServerDeployment

A prompt registered in one MCP server deployment. It is used to: - Create one prompt and register it in one MCP server deployment. - List tools registered in one MCP server deployment.

Variables:

id (str) – The identifier of prompt.
name (str) – The prompt name.
type (str) – The prompt type. It is a camelCase string representation of TypeOfPromptInUserMCPServerDeployment.
created_at (str) – Datetime when the prompt is created. It is formatted as RFC3339 UTC, e.g., 2026-02-24T19:12:48.285320Z.
user_id (bool) – The identifier of user who created the prompt.
user_name (str) – The name of user who created the prompt.
mcp_server_deployment_id (str) – The identifier of MCP server deployment (custom model deployment) under which the prompt is registered.

classmethod create(mcp_server_deployment_id, name, type)

Create a new MCP prompt and return it.

Parameters:

mcp_server_deployment_id (str) – The identifier of MCP server deployment (custom model deployment) under which the prompt is registered.
name (str) – The prompt name.
type (TypeOfPromptInUserMCPServerDeployment) – The prompt type.

Returns:

The created MCP prompt.

Return type:

PromptInUserMCPServerDeployment

classmethod list(mcp_server_deployment_id, offset=0, limit=10)

Get a list of MCP prompts.

Parameters:

mcp_server_deployment_id (str) – The identifier of MCP server deployment (custom model deployment) under which the prompt is registered.
offset (int) – The offset of the query.
limit (int) – The limit of returned MCP prompt.

Returns:

A list of MCP prompts.

Return type:

List[PromptInUserMCPServerDeployment]

delete()

Delete a MCP prompt.

Return type:: None

classmethod from_data(data)

Instantiate an object of this class using a dict.

Parameters:: data (dict) – Correctly snake_cased keys and their values.
Return type:: TypeVar(T, bound= APIObject)

classmethod from_server_data(data, keep_attrs=None)

Instantiate an object of this class using the data directly from the server, meaning that the keys may have the wrong camel casing

Parameters:

data (dict) – The directly translated dict of JSON from the server. No casing fixes have taken place
keep_attrs (iterable) – List, set or tuple of the dotted namespace notations for attributes to keep within the object structure even if their values are None

Return type:

TypeVar(T, bound= APIObject)

class datarobot._experimental.models.user_mcp_server_deployment.ResourceInUserMCPServerDeployment

A resource registered in one MCP server deployment. It is used to: - Create one resource and register it in one MCP server deployment. - List tools registered in one MCP server deployment.

Variables:

id (str) – The identifier of resource.
name (str) – The resource name.
type (str) – The resource type. It is a camelCase string representation of TypeOfResourceInUserMCPServerDeployment.
uri (str) – The resource URI.
created_at (str) – Datetime when the resource is created. It is formatted as RFC3339 UTC, e.g., 2026-02-24T19:12:48.285320Z.
user_id (bool) – The identifier of user who created the resource.
user_name (str) – The name of user who created the resource.
mcp_server_deployment_id (str) – The identifier of MCP server deployment (custom model deployment) under which the resource is registered.

classmethod create(mcp_server_deployment_id, name, type, uri)

Create a new MCP resource and return it.

Parameters:

mcp_server_deployment_id (str) – The identifier of MCP server deployment (custom model deployment) under which the resource is registered.
name (str) – The resource name.
type (TypeOfResourceInUserMCPServerDeployment) – The resource type.
uri (str) – The resource URI.

Returns:

The created MCP resource.

Return type:

ResourceInUserMCPServerDeployment

classmethod list(mcp_server_deployment_id, offset=0, limit=10)

Get a list of MCP resources.

Parameters:

mcp_server_deployment_id (str) – The identifier of MCP server deployment (custom model deployment) under which the resource is registered.
offset (int) – The offset of the query.
limit (int) – The limit of returned MCP resource.

Returns:

A list of MCP resources.

Return type:

List[ResourceInUserMCPServerDeployment]

delete()

Delete a MCP resource.

Return type:: None

classmethod from_data(data)

Instantiate an object of this class using a dict.

Parameters:: data (dict) – Correctly snake_cased keys and their values.
Return type:: TypeVar(T, bound= APIObject)

classmethod from_server_data(data, keep_attrs=None)

Instantiate an object of this class using the data directly from the server, meaning that the keys may have the wrong camel casing

Parameters:

data (dict) – The directly translated dict of JSON from the server. No casing fixes have taken place
keep_attrs (iterable) – List, set or tuple of the dotted namespace notations for attributes to keep within the object structure even if their values are None

Return type:

TypeVar(T, bound= APIObject)

class datarobot._experimental.models.user_mcp_server_deployment.ToolInUserMCPServerDeployment

A tool registered in one MCP server deployment. It is used to: - Create one tool and register it in one MCP server deployment. - List tools registered in one MCP server deployment.

Variables:

id (str) – The identifier of tool.
name (str) – The tool name.
type (str) – The tool type. It is a camelCase string representation of TypeOfToolInUserMCPServerDeployment.
created_at (str) – Datetime when the tool is created. It is formatted as RFC3339 UTC, e.g., 2026-02-24T19:12:48.285320Z.
user_id (bool) – The identifier of user who created the tool.
user_name (str) – The name of user who created the tool.
mcp_server_deployment_id (str) – The identifier of MCP server deployment (custom model deployment) under which the tool is registered.

classmethod create(mcp_server_deployment_id, name, type)

Create a new MCP tool and return it.

Parameters:

mcp_server_deployment_id (str) – The identifier of MCP server deployment (custom model deployment) under which the tool is registered.
name (str) – The tool name.
type (TypeOfToolInUserMCPServerDeployment) – The tool type.

Returns:

The created MCP tool.

Return type:

ToolInUserMCPServerDeployment

classmethod list(mcp_server_deployment_id, offset=0, limit=10)

Get a list of MCP tools.

Parameters:

mcp_server_deployment_id (str) – The identifier of MCP server deployment (custom model deployment) under which the tool is registered.
offset (int) – The offset of the query.
limit (int) – The limit of returned MCP tool.

Returns:

A list of MCP tools.

Return type:

List[ToolInUserMCPServerDeployment]

delete()

Delete a MCP tool.

Return type:: None

classmethod from_data(data)

Instantiate an object of this class using a dict.

Parameters:: data (dict) – Correctly snake_cased keys and their values.
Return type:: TypeVar(T, bound= APIObject)

classmethod from_server_data(data, keep_attrs=None)

Instantiate an object of this class using the data directly from the server, meaning that the keys may have the wrong camel casing

Parameters:

data (dict) – The directly translated dict of JSON from the server. No casing fixes have taken place
keep_attrs (iterable) – List, set or tuple of the dotted namespace notations for attributes to keep within the object structure even if their values are None

Return type:

TypeVar(T, bound= APIObject)

class datarobot._experimental.models.user_mcp_server_deployment.TypeOfPromptInUserMCPServerDeployment

Supported types of prompts in one user MCP server deployment

USER_PROMPT_TEMPLATE: A prompt template created as an MCP prompt decorated function within the user MCP server.
USER_PROMPT_TEMPLATE_VERSION: A prompt template created and registered in DataRobot.

class datarobot._experimental.models.user_mcp_server_deployment.TypeOfResourceInUserMCPServerDeployment

Supported types of resources in one user MCP server deployment

USER_RESOURCE: A resource created as an MCP resource decorated function within the user MCP server.

class datarobot._experimental.models.user_mcp_server_deployment.TypeOfToolInUserMCPServerDeployment

Supported types of tools in one user MCP server deployment

USER_TOOL: A tool created as an MCP tool decorated Python function within the user MCP server.
BUILT_IN_TOOL: A DataRobot Predictive AI tool or wrapper tool of external service (e.g., GitHub).
USER_TOOL_DEPLOYMENT: A tool created as a custom inference model.

class datarobot._experimental.models.user_mcp_server_version.PromptInUserMCPServerVersion

A prompt registered in one MCP server version. It is used to: - List prompts registered in one MCP server version.

Variables:

id (str) – The identifier of prompt.
name (str) – The prompt name.
type (str) – The prompt type. It is a camelized string representation of TypeOfPromptInUserMCPServerVersion.
created_at (str) – Datetime when the prompt is created. It is formatted as RFC3339 UTC, e.g. 2026-02-24T19:12:48.285320Z
user_id (bool) – The identifier of user who created the prompt.
user_name (str) – The name of user who created the prompt.
mcp_server_version_id (str) – The identifier of MCP server version (custom model version) under which the prompt is registered.

classmethod list(mcp_server_version_id, offset=0, limit=10)

Get a list of MCP prompts.

Parameters:

mcp_server_version_id (str) – The identifier of MCP server version (custom model version) under which the prompt is registered.
offset (int) – The offset of the query.
limit (int) – The limit of returned MCP prompt.

Returns:

A list of MCP prompts.

Return type:

List[PromptInUserMCPServerVersion]

classmethod from_data(data)

Instantiate an object of this class using a dict.

Parameters:: data (dict) – Correctly snake_cased keys and their values.
Return type:: TypeVar(T, bound= APIObject)

classmethod from_server_data(data, keep_attrs=None)

Instantiate an object of this class using the data directly from the server, meaning that the keys may have the wrong camel casing

Parameters:

data (dict) – The directly translated dict of JSON from the server. No casing fixes have taken place
keep_attrs (iterable) – List, set or tuple of the dotted namespace notations for attributes to keep within the object structure even if their values are None

Return type:

TypeVar(T, bound= APIObject)

class datarobot._experimental.models.user_mcp_server_version.ResourceInUserMCPServerVersion

A resource registered in one MCP server version. It is used to: - List resources registered in one MCP server version.

Variables:

id (str) – The identifier of resource.
name (str) – The resource name.
type (str) – The resource type. It is a camelized string representation of TypeOfResourceInUserMCPServerVersion.
uri (str) – The resource URI.
created_at (str) – Datetime when the resource is created. It is formatted as RFC3339 UTC, e.g. 2026-02-24T19:12:48.285320Z
user_id (bool) – The identifier of user who created the resource.
user_name (str) – The name of user who created the resource.
mcp_server_version_id (str) – The identifier of MCP server version (custom model version) under which the resource is registered.

classmethod list(mcp_server_version_id, offset=0, limit=10)

Get a list of MCP resources.

Parameters:

mcp_server_version_id (str) – The identifier of MCP server version (custom model version) under which the resource is registered.
offset (int) – The offset of the query.
limit (int) – The limit of returned MCP resource.

Returns:

A list of MCP resources.

Return type:

List[ResourceInUserMCPServerVersion]

classmethod from_data(data)

Instantiate an object of this class using a dict.

Parameters:: data (dict) – Correctly snake_cased keys and their values.
Return type:: TypeVar(T, bound= APIObject)

classmethod from_server_data(data, keep_attrs=None)

Instantiate an object of this class using the data directly from the server, meaning that the keys may have the wrong camel casing

Parameters:

data (dict) – The directly translated dict of JSON from the server. No casing fixes have taken place
keep_attrs (iterable) – List, set or tuple of the dotted namespace notations for attributes to keep within the object structure even if their values are None

Return type:

TypeVar(T, bound= APIObject)

class datarobot._experimental.models.user_mcp_server_version.ToolInUserMCPServerVersion

A tool registered in one MCP server version. It is used to: - List tools registered in one MCP server version.

Variables:

id (str) – The identifier of tool.
name (str) – The tool name.
type (str) – The tool type. It is a camelized string representation of TypeOfToolInUserMCPServerVersion
created_at (str) – Datetime when the tool is created. It is formatted as RFC3339 UTC, e.g. 2026-02-24T19:12:48.285320Z
user_id (bool) – The identifier of user who created the tool.
user_name (str) – The name of user who created the tool.
mcp_server_version_id (str) – The identifier of MCP server version (custom model version) under which the tool is registered.

classmethod list(mcp_server_version_id, offset=0, limit=10)

Get a list of MCP tools.

Parameters:

mcp_server_version_id (str) – The identifier of MCP server version (custom model version) under which the tool is registered.
offset (int) – The offset of the query.
limit (int) – The limit of returned MCP tool.

Returns:

A list of MCP tools.

Return type:

List[ToolInUserMCPServerVersion]

classmethod from_data(data)

Instantiate an object of this class using a dict.

Parameters:: data (dict) – Correctly snake_cased keys and their values.
Return type:: TypeVar(T, bound= APIObject)

classmethod from_server_data(data, keep_attrs=None)

Instantiate an object of this class using the data directly from the server, meaning that the keys may have the wrong camel casing

Parameters:

data (dict) – The directly translated dict of JSON from the server. No casing fixes have taken place
keep_attrs (iterable) – List, set or tuple of the dotted namespace notations for attributes to keep within the object structure even if their values are None

Return type:

TypeVar(T, bound= APIObject)

class datarobot._experimental.models.user_mcp_server_version.TypeOfPromptInUserMCPServerVersion

Supported types of prompts associated with one user MCP server version

USER_PROMPT_TEMPLATE: A prompt template created as a mcp prompt decorated function within the user MCP server.

class datarobot._experimental.models.user_mcp_server_version.TypeOfResourceInUserMCPServerVersion

Supported types of resources associated with one user MCP server version

USER_RESOURCE: A resource created as a mcp resource decorated function within the user MCP server.

class datarobot._experimental.models.user_mcp_server_version.TypeOfToolInUserMCPServerVersion

Supported types of tools associated with one user MCP server version

USER_TOOL: A tool created as a mcp tool decorated python function within the user MCP server.
BUILT_IN_TOOL: A DataRobot Predictive AI tool or wrapper tool of external service (e.g., github).