gemini-cli - 💡(How to fix) Fix Add Amazon URL parsing and metadata extraction for product comparisons [1 pull requests]

What would you like to be added?

🚀 Feature Description

Add a URL unfurling and scraping mechanism specifically for standard e-commerce links (like Amazon amzn.in short URLs) so that the AI can automatically extract product titles, pricing, and key specifications to answer comparison and analysis requests seamlessly.

❓ Problem Statement / Motivation

Currently, when a user drops shortened Amazon links (e.g., https://amzn.in/d/00geRr5g) into the chat and asks for a comparison, the AI is unable to resolve or scrape the contents of those links. This forces the user to manually copy-paste dense spec sheets, disrupting the user experience and breaking the natural flow of the conversation.

💡 Proposed Solution

1. URL Expander: Implement a safe URL resolver that detects shortened domains (amzn.to, amzn.in/d/, etc.) and expands them to their full canonical URLs.
2. E-Commerce Scraper/API Integration: Create a parser service that reads the expanded product pages to reliably extract:
   • Product Name / Model
   • Price
   • Top features/Specifications (CPU, RAM, Storage, Screen size for electronics)
3. Context Injection: Inject this extracted structural metadata into the LLM's prompt context so it can fulfill the request (e.g., "Compare both laptops") immediately.

🛠️ Alternative Solutions Considered

• User Prompt Guidance: Politely ask the user to provide the specifications manually (Current behavior, but introduces friction).
• Browsing Tool fallback: Rely on a generic web search tool, though this frequently fails or gets blocked by e-commerce anti-bot measures (like CAPTCHAs) without a specialized scraper/header configuration.

📝 Additional Context / Implementation Steps

• Create a regex utility to detect short Amazon URLs in user inputs.
• Set up an asynchronous request handler to expand shortened links and handle redirects.
• Build a robust parsing module (using beautifulsoup4, Playwright, or an official product API if available) that targets product title and bullet-point specs.
• Handle edge cases gracefully (e.g., product out of stock, regional blocks, or broken links) by falling back to a specific error message.
• Write integration tests simulating link inputs and ensuring the AI correctly receives the parsed context.

Why is this needed?

<img width="1033" height="841" alt="Image" src="https://github.com/user-attachments/assets/513fd2e2-c39c-4dde-bf7b-f59cbf6dad5c" />

Additional context

No response