Why not open the Automatic prompt generation for the agent chat?

[jaruan]

I noticed that automatic prompt generation doesn't support the agent chat. Could something affect the output?
code

[dosubot[bot]]

Hey @jaruan! I'm here to help you with any bugs, questions, or contributions you have. Let's tackle this together!

Automatic prompt generation for agent chat is influenced by specific conditions and configurations in the codebase. The relevant conditions include the strategy in agent_mode and the model_mode. If the strategy is not 'react_router' or 'router', and depending on whether the model_mode is 'completion' or 'chat', different prompt templates are used ^[1][2].

Additionally, the useAdvancedPromptConfig hook determines the configuration of prompts based on the promptMode and modelModeType parameters. If promptMode is set to PromptMode.advanced, the prompts are configured differently based on whether modelModeType is ModelModeType.chat or ModelModeType.completion ^[2].

There are also known limitations related to token calculation. If the remaining number of tokens is insufficient according to the model's context token size limit and max token size limit, an exception is raised ^[3].

These conditions and limitations could affect the output and functionality of automatic prompt generation for agent chat.

To continue talking to Dosu, mention @dosu.