The ``catalog`` package
===========================

.. py:module:: ansys.fluent.mcp.solve.catalog


Summary
-------








.. py:currentmodule:: catalog
.. tab-set::


    .. tab-item:: Submodules

        .. list-table::
          :header-rows: 0
          :widths: auto


          * - :py:obj:`~ansys.fluent.mcp.solve.catalog.help`
            - Docstring extractor for PyFluent's generated ``settings_<version>.py``.


          * - :py:obj:`~ansys.fluent.mcp.solve.catalog.index`
            - Lexical fallback over the PyFluent API catalog (the default in practice).


          * - :py:obj:`~ansys.fluent.mcp.solve.catalog.retriever`
            - API retrieval (RAG over the Fluent API catalog).


          * - :py:obj:`~ansys.fluent.mcp.solve.catalog.schema`
            - Static index over PyFluent's canonical ``settings.json`` schema.












.. toctree::
   :titlesonly:
   :maxdepth: 1
   :hidden:

    <span class="nf nf-fa-file"></span> help</api/ansys/fluent/mcp/solve/catalog/help/index>
    <span class="nf nf-fa-file"></span> index</api/ansys/fluent/mcp/solve/catalog/index/index>
    <span class="nf nf-fa-file"></span> retriever</api/ansys/fluent/mcp/solve/catalog/retriever/index>
    <span class="nf nf-fa-file"></span> schema</api/ansys/fluent/mcp/solve/catalog/schema/index>











Description
-----------


Solve settings catalog (discovery, indexing, retrieval).

* :mod:`ansys.fluent.mcp.solve.catalog.schema`: Offline-cached
  PyFluent settings tree (~62k paths) used by the plan validator and
  the indexer.
* :mod:`ansys.fluent.mcp.solve.catalog.index`: Keyword inverted
  index built on top of the schema. Drives ``find_setting_path``.
* :mod:`ansys.fluent.mcp.solve.catalog.help`: Help/description
  dataclasses attached to catalog entries.
* :mod:`ansys.fluent.mcp.solve.catalog.retriever`: Semantic plus
  lexical retriever that returns ranked candidates for
  ``find_api``/RAG flows.

The catalog is consumed by BOTH backends (path validation,
``find_setting_path``) and the agent intelligence layer (RAG
grounding). Keeping it separate makes that dual ownership explicit.















..
    !! processed by numpydoc !!