Baehn
Prompting Gpt5 (with thinking), somethimes results in the following error: `Error executing callback: .../.vim/plugged/gp.nvim/lua/gp/dispatcher.lua:373: Expected value but found invalid token at character 1 stack traceback: [C]: in function 'decode' .../.vim/plugged/gp.nvim/lua/gp/dispatcher.lua:373: in function 'out_reader' .../.vim/plugged/gp.nvim/lua/gp/tasker.lua:166: in function </home/chbaeh/.vim/plugged/gp.nvim/lua/gp/tasker.lua:158> ` - I never expierenced this with the pure chat model (gpt-5-chat-latest) or other models. - Using the same prompts with curl works. - I only occurs with complex prompts, i.e. when the model thinks alot before responding. Maybe it is a timeout problem? - my nvim version is 0.11.3 - installed gp.nvim is on commit c37f154b97690c4925fef4e35ffdbf2c844b5f4e