github DearVa/Everywhere v0.6.7
Everywhere v0.6.7

4 hours ago

✨ Features

  • Visual Context Supercharged: Introducing get_visual_tree tool. Your assistant can now call this tool to fetch on-screen elements freely, massively boosting its contextual comprehension and self-awareness capabilities. Note: This is an experimental feature and may not work perfectly with all secenarios. Feedback is highly appreciated!
  • Model Update: Removed the deprecated Gemini 3 Pro Preview model and added the newly released Gemini 3.1 Pro Preview.

🚀 Improvements

  • Framework Upgrade: Updated the underlying MCP (Model Context Protocol) framework version, which may resolve previously reported edge-case issues (#232).

🐛 Bug Fixes

  • Critical Context Fix (Windows): Fixed a fatal typo in the visual context extraction algorithm on Windows. This bug previously caused premature truncation of complex contexts, hiding crucial information below the selected element from the assistant.
  • Resolved a rendering crash that could occur after enabling the text selection feature (#313).
  • Fixed a random crash bug related to Grid layout calculations, bringing further stability to the application.

Full Changelog: v0.6.6...v0.6.7

Don't miss a new Everywhere release

NewReleases is sending notifications on new releases.