Say Hello to Vision Assistant: Your Instant, Shortcut-Triggered Desktop AI Tool

``This is a submission for the Sanity Challenge, Path Two: Vibe-Code Something Strange

What I Built

I built Vision Assistant v1.0.0, a custom desktop AI vision tool built from scratch using Electron. It is designed to bring instant, system-wide AI assistance directly to the desktop environment without the friction of constant context-switching and this version package is built only for windows user right now.

** Key Features: **

Global Shortcut Activation:
Runs in the background and triggers instantly from any active window or application using the Ctrl + Shift + Space shortcut.

Real-Time Chat & Vision:
Seamlessly process and analyze visual content using integrated AI capabilities right from the desktop.

Multiple Screen Capture:
Allows capturing and analyzing multiple screens or active display areas simultaneously, perfect for multi-monitor workflows.

Native Windows Experience:

Lightweight standalone Windows application packaged via electron-builder with a sleek dark/purple UI theme. )

Demo

Setup Guide & Walkthrough Video:
[https://drive.google.com/drive/folders/1Kci00sPzQalZA78Urge4v8mkXLHp4dbR?usp=sharing]

Explore the repository and release notes on GitHub:
GitHub Release Page

Grab the direct package:
Direct Download Link

You can download the latest Windows release bundle directly from the GitHub repository releases.

Code

GitHub Repository:

https://github.com/chhavigupta783/vission-assisstant-v1

My Build Process

AI-Native Setup:
Leveraged AI capabilities to vibe-code and structure the entire Electron desktop app from scratch, handling native window bindings and global shortcuts.

AI Integration:
Powered by the Gemini API to handle real-time chat, vision analysis, and contextual interactions seamlessly.

Challenges Faced:
Managing local environment configurations, ensuring smooth packaging with electron-builder, and handling window focus events for system-wide shortcuts.

Course-Correction:

Refined the build setup documentation to guide users properly through extracting zip assets and bypassing Windows SmartScreen warnings for custom unsigned executables.

Sanity Project Details

Note: While the core application is built natively using Electron and the Gemini API, Sanity's documentation was referenced for content modeling patterns and architecture guidelines.

[A sleek, professional 3D graphic featuring a vibrant purple glowing bubble that showcases the dark-mode UI of the Vision Assistant desktop application. At the bottom, a stylish UI element clearly highlights the core system-wide feature with the text "press ctrl + shift + space".]

(https://dev-to-uploads.s3.us-east-2.amazonaws.com/uploads/articles/0ja24tx9sglq2sco303q.png)

---

<!-- Thank you for visiting --!>