Can I vibecode Genei?

Build a robust, deliberately narrow personal substitute for Genei, an AI-powered research and summarization workspace. This is not a full clone, but it must capture the core value proposition: drastically reducing the time spent reading and synthesizing dense, long-form content.

Use exactly this tech stack: Next.js 15 (App Router), TypeScript, TailwindCSS for styling, SQLite for local storage, and Playwright for E2E testing. Use OpenAI API or Anthropic API for the summarization capabilities.

Primary Objective:
Build a private, single-user research summarization workspace that allows importing user-supplied URLs or text files, extracting metadata, generating AI-powered summaries, supporting rich-text notes, and searching the local document corpus.

Core Features & UI Layout:
1. Dashboard & Document Library:
   - A sidebar navigation listing all imported documents, folders/projects, and tags.
   - A main view displaying a grid or list of documents with metadata (title, date added, source URL/file name, status).
   - "Add Document" floating action button or prominent header button.
2. Document Viewer & Summarizer (The Core Loop):
   - A split-pane view. Left pane: The original text content of the imported document. Right pane: The AI research assistant interface.
   - Auto-generated Summaries: Break down the document into sections, key bullet points, and extract top keywords based on frequency and relevance.
   - Search & Question Mode: A chat-like input in the right pane where users can ask specific questions about the document, and the AI answers strictly based on the text.
3. Integrated Research Workflow:
   - Built-in Notepad: A section within the right pane or a dedicated tab for users to take notes, annotate, and organize ideas alongside the document.
   - Text selection actions: Highlight text in the left pane to instantly append it to notes, summarize just that section, or find related content.

Step-by-Step Implementation Instructions:
- Start from an empty folder and scaffold the Next.js project.
- Implement the SQLite database schema for Documents, Notes, and Tags.
- Build the import functionality (accepting raw text or basic URL scraping) and store the content locally.
- Integrate the AI API (using a mocked response first, then requiring an API key via `.env`). Put every secret and external credential in `.env` and provide `.env.example`.
- Design the split-pane UI using Tailwind, focusing on a premium, modern aesthetic with smooth transitions and accessible contrast.
- Implement the "Question Mode" feature using the AI API to query the document content.
- Ensure all user data is stored entirely locally in SQLite.
- Include clear empty, loading, validation, success, and failure states throughout the application.
- Add import/export functionality for all data to prevent vendor lock-in.

Exclusions & Constraints:
- Do NOT add analytics, telemetry, ads, or third-party accounts.
- Deliberately exclude paid-product advantages like licensed scholarly metadata, complex citation graph scale, or publisher-specific import reliability.
- Write focused unit tests for the data model and one end-to-end smoke test that proves the core document import and summarization loop works.
- Create a comprehensive README with setup, architecture, and limitations.