FluidsLeafMCP

class ansys.cfx.mcp.common.base.FluidsLeafMCP(*, backends: dict[str, ansys.cfx.mcp.common.backend.Backend], expose_tools: Iterable[str] = ALL_TOOLS, hide_connection_tools: bool = False, conversation_store: ansys.cfx.mcp.common.conversation.ConversationStore | None = None, codegen_pipeline: ansys.cfx.mcp.common.codegen.CodegenPipeline | None = None, name: str | None = None, **fastmcp_kwargs: Any)

Bases: ansys.common.mcp.server.PyAnsysBaseMCP

Base server class for a single shared MCP leaf.

Overview

Methods

run	Run the MCP server event loop for the configured transport.
product_startup	Run product-specific startup hooks before serving MCP requests.
product_cleanup	Run product-specific cleanup hooks before shutdown.
build_toolsets	Build toolset metadata for the tools exposed by this leaf server.
register_run_code_observer	Register a callback that observes successful run-code executions.

Properties

backend

Return the currently active backend instance.

Attributes

leaf_name
default_backend_kind
cache_ttl_seconds
connect_on_startup
component_label
error_remediation_description

Import detail

from ansys.cfx.mcp.common.base import FluidsLeafMCP

Property detail

property FluidsLeafMCP.backend: ansys.cfx.mcp.common.backend.Backend

Return the currently active backend instance.

Returns:

Backend

Value computed by the helper for the requested CFX workflow.

Attribute detail

FluidsLeafMCP.leaf_name: str = 'fluids'

FluidsLeafMCP.default_backend_kind: str | None = None

FluidsLeafMCP.cache_ttl_seconds: float = 30.0

FluidsLeafMCP.connect_on_startup: bool = False

FluidsLeafMCP.component_label: str = ''

FluidsLeafMCP.error_remediation_description: str = 'Generate a Markdown remediation / how-to answer for a natural-language request (error message,...

Method detail

FluidsLeafMCP.run(transport: str = 'stdio', host: str | None = None, port: int | None = None) → None

Run the MCP server event loop for the configured transport.

Parameters:

transportstr, optional

LLM transport implementation to use. Default is 'stdio'.

hoststr | None, optional

Host interface for the MCP transport. Default is None.

portint | None, optional

Port of the running CFX service to attach to. Default is None.

Returns:

None

No value is returned; side effects are applied to the relevant cache, session, or server.

FluidsLeafMCP.product_startup() → None

Run product-specific startup hooks before serving MCP requests.

Returns:

None

No value is returned; side effects are applied to the relevant cache, session, or server.

FluidsLeafMCP.product_cleanup() → None

Run product-specific cleanup hooks before shutdown.

Returns:

None

No value is returned; side effects are applied to the relevant cache, session, or server.

FluidsLeafMCP.build_toolsets() → list[dict[str, Any]]

Build toolset metadata for the tools exposed by this leaf server.

Returns:

list[dict[str, Any]]

List of structured records matching the request.

FluidsLeafMCP.register_run_code_observer(observer: Callable[Ellipsis, Awaitable[None] | None]) → None

Register a callback that observes successful run-code executions.

Parameters:

observerCallable[…, Awaitable[None] | None]

Callback to notify after run-code execution.

Returns:

None

No value is returned; the callback is stored for later notifications.