papers_rag.app
==============

.. py:module:: papers_rag.app

.. autoapi-nested-parse::

   Streamlit query chat over the local papers Chroma index.



Attributes
----------

.. autoapisummary::

   papers_rag.app.MESSAGES_STATE_KEY
   papers_rag.app.STRATEGY_STATE_KEY


Classes
-------

.. autoapisummary::

   papers_rag.app.ChatRole
   papers_rag.app.ChatMessageKey


Functions
---------

.. autoapisummary::

   papers_rag.app.main
   papers_rag.app.run


Module Contents
---------------

.. py:class:: ChatRole

   Bases: :py:obj:`enum.StrEnum`


   Roles stored in the Streamlit chat transcript.


   .. py:attribute:: USER
      :value: 'user'



   .. py:attribute:: ASSISTANT
      :value: 'assistant'



.. py:class:: ChatMessageKey

   Bases: :py:obj:`enum.StrEnum`


   Keys on each message dict in ``st.session_state``.


   .. py:attribute:: ROLE
      :value: 'role'



   .. py:attribute:: CONTENT
      :value: 'content'



   .. py:attribute:: CITATIONS
      :value: 'citations'



.. py:data:: MESSAGES_STATE_KEY
   :type:  Final
   :value: 'messages'


.. py:data:: STRATEGY_STATE_KEY
   :type:  Final
   :value: 'rag_strategy'


.. py:function:: main() -> None

   Render the Papers RAG chat UI.


.. py:function:: run() -> None

   Launch Streamlit for the papers chat (console script ``papers-rag``).

   Extra CLI args after ``papers-rag`` are forwarded to ``streamlit run``.