Description
Enable the embedded chat widget to gain "vision" into the parent website by extracting page context (headers, titles, content, metadata) and providing it to the LLM for more contextual and relevant responses. This enhances the chat widget's understanding of the user's browsing environment.
Key Features:
- PostMessage API communication (iframe ↔ parent)
- Basic page information extraction (URL, title, headers)
- Advanced context extraction (meta tags, structured data)
- Secure context transmission
- Backend context processing
Impact:
- More contextual responses based on page content
- Better understanding of user's browsing context
- Improved user experience for embedded widgets
- Enhanced relevance of LLM responses
Reward
🎉 180,000 GURU tokens
Getting Started
- Review design document:
bot/docs/feature/iframe-context-extraction.md
- Check chat widget implementation in frontend
- Review WebSocket message handling:
ag_ui_gateway/
- Explore PostMessage API usage
Implementation Phases
Phase 1: Basic Context Extraction (90,000 GURU)
- PostMessage communication setup
- Basic page information extraction
- Context transmission with chat messages
Phase 2: Advanced Context (90,000 GURU)
- Meta information extraction
- Structured data parsing (JSON-LD, microdata)
- User interaction context
Acceptance Criteria
Description
Enable the embedded chat widget to gain "vision" into the parent website by extracting page context (headers, titles, content, metadata) and providing it to the LLM for more contextual and relevant responses. This enhances the chat widget's understanding of the user's browsing environment.
Key Features:
Impact:
Reward
🎉 180,000 GURU tokens
Getting Started
bot/docs/feature/iframe-context-extraction.mdag_ui_gateway/Implementation Phases
Phase 1: Basic Context Extraction (90,000 GURU)
Phase 2: Advanced Context (90,000 GURU)
Acceptance Criteria