allow AI Assistance to run against any chat-completion-style endpoint (#630)

This adds an ASSISTANT_CHAT_COMPLETION_ENDPOINT which can be used
to enable AI Assistance instead of an OpenAI API key. The assistant
then works against compatible endpoints, in the mechanical sense.
Quality of course will depend on the model. I found some tweaks
to the prompt that work well both for Llama-2 and for OpenAI's models,
but I'm not including them here because they would conflict with some
prompt changes that are already in the works.

Co-authored-by: Alex Hall <alex.mojaki@gmail.com>

app/client/widgets/FormulaAssistant.ts

@@ -6,7 +6,7 @@ import {movable} from 'app/client/lib/popupUtils';
 import {logTelemetryEvent} from 'app/client/lib/telemetry';
 import {ColumnRec, ViewFieldRec} from 'app/client/models/DocModel';
 import {ChatMessage} from 'app/client/models/entities/ColumnRec';
-import {HAS_FORMULA_ASSISTANT} from 'app/client/models/features';
+import {HAS_FORMULA_ASSISTANT, WHICH_FORMULA_ASSISTANT} from 'app/client/models/features';
 import {getLoginOrSignupUrl, urlState} from 'app/client/models/gristUrlState';
 import {buildHighlightedCode} from 'app/client/ui/CodeHighlight';
 import {autoGrow} from 'app/client/ui/forms';
@@ -879,7 +879,7 @@ class ChatHistory extends Disposable {
               '"Please calculate the total invoice amount."'
             ),
           ),
-          cssAiMessageBullet(
+          (WHICH_FORMULA_ASSISTANT() === 'OpenAI') ? cssAiMessageBullet(
             cssTickIcon('Tick'),
             dom('div',
               t(
@@ -891,7 +891,7 @@ class ChatHistory extends Disposable {
                 }
               ),
             ),
-          ),
+          ) : null,
         ),
         cssAiMessageParagraph(
           t(