Can I vibecode Glasp?

Build a comprehensive, modern web application that serves as a personal replacement for Glasp, a social web highlighter and knowledge management platform.
The application should use Next.js 15 (App Router), TypeScript, Tailwind CSS (for styling, using a premium, vibrant aesthetic with micro-animations), and SQLite for local data storage. 

Core Objectives:
Create a dynamic web highlighting and knowledge organization workspace. The app must allow users to save URLs or upload PDFs, extract metadata (title, author, date, cover image), highlight specific text within the content, and attach personal notes to those highlights.

Feature Specifications:
1. Dashboard/Home:
- A visually stunning dashboard displaying recent highlights, saved articles, and a "knowledge graph" visualization of tags and topics.
- Use a glassmorphic design language with smooth gradients and a polished dark mode option.
- Sidebar for easy navigation between: Home, My Library (saved items), Highlights, Tags, and AI Assistant.

2. Content Ingestion & Highlighting:
- A mechanism to add new content via URL input or file upload (PDFs, markdown).
- An interactive reader view where users can select text to highlight. Implement a floating toolbar that appears upon text selection, offering different highlight colors (yellow, green, pink, blue) and a "Add Note" button.
- Extract and display reading time and a generated summary (simulate AI summarization if no real API is used, or optionally integrate OpenAI API via `.env`).

3. Knowledge Organization:
- Support for tagging and categorizing highlights.
- A robust search functionality to query the local SQLite corpus across titles, content, notes, and tags.

4. "AI Clone" / Chat Interface:
- A dedicated chat interface mimicking Glasp's AI Clone feature. Users can "chat" with their own highlights. 
- Implement a simulated streaming response interface where the "AI" references the user's saved notes. If an OpenAI key is provided in `.env`, use it to answer questions grounded in the local SQLite database.

5. Social/Export capabilities:
- While primarily single-user, add a "Public Profile" preview mode to simulate the social aspect, allowing users to generate a shareable link or export their highlights to markdown (for Obsidian/Notion).

Step-by-Step Instructions for the AI Builder:
1. Initialize the Next.js project with Tailwind CSS and install necessary dependencies (e.g., Lucide React for icons, Framer Motion for animations).
2. Set up the local SQLite database schema for `Articles`, `Highlights`, `Notes`, and `Tags`.
3. Build the core layout, sidebar navigation, and the premium glassmorphic UI shell.
4. Implement the 'Add URL' feature: fetch page metadata (mocked if necessary) and display it in the library.
5. Build the Reader View with the text selection and highlighting logic.
6. Create the Highlights dashboard and the search functionality.
7. Implement the 'AI Clone' chat interface.
8. Ensure all loading states, empty states, and errors are handled gracefully with polished micro-animations.
9. Write a comprehensive README with setup instructions.

Do not include third-party analytics or telemetry. Ensure keyboard accessibility and high contrast.