Sven Olderaan 247abbe91a Improve function registration for SillyBubble extension
- Use the proper ToolManager API when available
- Add fallback mechanisms when ToolManager is not available
- Fix extension settings handling to use the standard pattern
- Implement proper JSON schema for function parameters
- Add delayed initialization to ensure SillyTavern is fully loaded
- Add better error handling and logging

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>
2025-03-14 21:20:23 +01:00

SillyBubble - Dynamic Chat Bubble Extension for SillyTavern

SillyBubble is a SillyTavern extension that enables AI models to generate dynamic chat bubble images with text. The extension provides a function that AI models can call to create properly formatted markdown image links with URL-encoded text parameters.

Features

  • Provides a function for AI to generate markdown image links
  • Properly URL-encodes text parameters for image generation
  • Supports different bubble styles through style parameters
  • Configurable image service URL
  • Simple testing interface to preview bubble generation

Installation and Usage

Installation

  1. Install the extension using SillyTavern's built-in extension installer
  2. Configure the image service URL in the extension settings

Usage

  1. The extension registers a custom function generateChatBubbleImage that AI models can call
  2. When called, it returns a properly formatted markdown image link:
    ![](image.php?q=Hello%20world)
    
  3. The function accepts two parameters:
    • text: The text to display in the bubble (required)
    • style: The visual style of the bubble (optional)

Prerequisites

  • SillyTavern with custom function support
  • A properly configured image generation service that accepts text parameters

Support and Contributions

For support or to contribute to this project, please visit the GitHub repository or open an issue.

License

MIT License

Description
No description provided
Readme 163 KiB
Languages
JavaScript 38.6%
PHP 30.2%
HTML 28.1%
CSS 3.1%