Release Notes - Version 26.4.27: Mentions

This update introduces several major features and quality-of-life improvements to make your X-Ray experience smoother, faster, and more localized.

New Features & Enhancements

• Entity Mentions & Navigation
  You can now see exactly where characters and locations appear throughout your book. The new "Mentions" feature provides a list of occurrences, allowing you to jump directly to that specific page. We've also polished the navigation so that menus close automatically when you jump, getting you back to your reading immediately.

• Improved GPT & AI Compatibility
  We've standardized the way the plugin communicates with OpenAI models. This update includes better handling for newer models like GPT-4o, improved token management to prevent errors, and a more robust fallback system to ensure your requests always go through.

• Smarter Background Fetching
  The "inline fetching" logic has been overhauled to be less intrusive. Data for new chapters is now merged more efficiently in the background, significantly reducing the "freezing" or lag that could occur on older hardware like the Kindle Paperwhite when the plugin was busy fetching new insights.

• Performance & Stability
  Under the hood, we've completed a major refactoring of the plugin's architecture. This makes the code faster, more reliable, and easier for us to maintain, ensuring a more stable experience on all devices.

Bug Fixes

• Resolved potential crashes related to multi-byte character handling in AI responses.
• Fixed inconsistent token limit settings for various AI models.

Full Changelog: 26.4.26...26.4.27