aidial_assistant/application/assistant_application.py (3 lines):
	- line 200: # TODO: Add max_addons_dialogue_tokens as a request parameter
	- line 245: # TODO: else compare the history size to the max prompt tokens of the underlying model
	- line 280: # TODO: Add max_addons_dialogue_tokens as a request parameter


aidial_assistant/model/model_client.py (2 lines):
	- line 127: # TODO: Use a dedicated endpoint for counting tokens.
	- line 150: # TODO: Use a dedicated endpoint for discarded_messages.


aidial_assistant/chain/command_chain.py (1 line):
	- line 128: # TODO: Limit the error message size. The error message should not exceed reserved assistant overheads.