Report whether the cfx leaf has an active backend. Safe to call before connect. Returns the connected endpoint, backend kind, and the list of tools available in the current state.

Connect the cfx leaf to a backend. Available backend kinds: ['pycfx']. Pass backend_kind to choose, or omit it to auto-select. Backend-specific options (url/token/ip/port/...) go in connect_kwargs as a dict and are forwarded to the backend.

Disconnect the cfx leaf's active backend.

Generate Python code for the cfx stage from a natural-language prompt. Returns either status='ok' with code, or status='needs_clarification' with one or more clarifications to answer via the clarify tool. Pass the returned session_id back on follow-up calls to keep context.

Provide an answer to a clarification raised by codegen. session_id and clarification_id come from the previous codegen (or clarify) response.

Execute Python code against the active PyCFX session namespace. The code runs with pre, solver, post, session, cfxpre, cfxsolver, and cfxpost helpers refreshed from the current CFX sessions. Returns stdout, stderr, and any __return__ value. Prefer cfx_workflow or cfx_model_context for routed actions and read-only queries.

Dry-run / validate the generated code without applying side effects. Returns parse / type / semantic feedback for the LLM.

Run one focused CFX lifecycle or artifact action. Actions: start_pre, import_mesh, write_def, start_solver, wait_solver, get_results_file, open_post, status. Use codegen for custom model setup edits.

Return a targeted, compact CFX model context slice. Actions: summary, list_named_objects, find_named_object, select_named_objects, state, api_help, find_api, allowed_values, targeted_context. Use max_items to keep responses small.