• 

  hey-nano-side-panel

• 

  Voice Mode - Input Prompt

Hey Nano

Inspiration

• AI assistants often misread web context.
  
• The goal: let users show the AI what matters instead of relying on guesswork.
  

What it does

• Adds a draggable context pane to Chrome.
  
• Lets users highlight or capture webpage sections to feed curated context directly to the LLM.
  

How we built it

• Chrome extension built with HTML, CSS, and JS.
  
• Background scripts manage mic state and context syncing.
  
• Sidebar handles chat and visual context feedback.
  

Challenges we ran into

• Maintaining tab and context state across multiple windows.
  
• Balancing simplicity with DOM inspection tools.
  
• At this stage, images are not recognized well — a computer vision module would perform better for next iterations.
  

Accomplishments that we're proud of

• Clean, responsive UI that keeps the AI focused only on chosen content.
  
• Stable context handling and easy-to-use visual interface.
  

What we learned

• Voice control is secondary to context clarity.
  
• Users prefer precision over automation.
  
• Chrome’s side panel allows deeper integrations than expected.
  

What's next for Hey Nano

• Add fine-grained context sharing between tabs.
  
• Integrate computer vision for image-based context.
  
• Voice mode roadmap open to add more voice commands.
  

Built With

• api
• built-in
• ccs
• gemini
• html
• javascript
• nano
• prompt
• speech
• user-popup.js-(line-31)
• vainilla
• web