Bump libraries and use gpt-4o-mini as default model

Fix some compatibility issues with spigot

Fix missing relocation for bstats

Let people choose if they want to share data or not

Implement bstats

Let it do the resume even when an autoresume was triggered
Do not display the bossbar (fix npe)

Zoom in when talking to npc

Let people select if they want to get short answers from the character

Fix wrong lang key

Let characters have knowledge about their location via character-location config option

Let people use tags to start a conversation. Ported from premium version.

Add sounds to main commands

Let people enable a bossbar to notify the user when in conversation

Do not let people apply more than 1 character at a time. (trait)

Let people set the lang as “auto_detect” so the plugin will use the client language.

Allow rgb in lang file.

Add "my_limits" as a placeholder to display both storage and interaction.

Add player interactions (rem_interactions) placeholder

Add default_interactions and default_storage placeholder

Add prefix and lang placeholders

Add role and personality placeholder

Add id placeholder

Add name placeholder
%gpttalk_<characterId>_name%
Description: Returns the name of the character based on their ID.

Let people use /gpttalk start <id> command

Some code refactor & minimal performance improvements for future changes.

LINK FIX

Allow people set the conversation lang.

Don't allow duplicated characters when loading them.

Move plugin default characters to "defaultcharacters" folder.

Added "enable" option at characters config to decide whether it should be loaded.

Make the "characters" folder fully available for user characters only.

Fixed missing lib.

LINK FIX

Added update checker so you get a notification when we update the plugin!

Added more info at config about how to pick the models for each chat mode.

Let people use the debug option to easily catch unexpected issues.

Trim not needed whitespaces added from OpenAI.

Bump max tokens to 600

Code improvements and other versions reminder.

Auto update the lang file
Hotfix. Missing libs when loading it on paper

Improve prompts, add "wiseman", and allow the setup of the character's personality.

WARNING: You have to remove the harry potter config to include the changes in the others that have since the previous version.

Forcefully quit the conversation if some exception is triggered.

Improved the backup and auto-resume system to ensure it doesn't interfere with ongoing chat history.

Allowed the AI to say goodbye using a custom prompt when the limit is reached or when the 'bye' keyword is triggered.

Changed the 'cancel' keyword to 'bye' keyword.

Remove not needed shaded lib.
Improved resume prompt to cover chat and completion cases.

It is now allowed to use the npc's name in commands instead of just the id

It is now allowed to use the npc's name in commands instead of just the id

Refined so that the gpt 3.5 model can be used in chat mode without breaking character by adding a suffix. The default model will now be gpt 3.5 and chat mode, this model spends considerably less tokens.

Stop phrase is used in chat.
Default chat is used.
Suffix added to avoid out of context.

Fixed AI speaking for you in completion mode