Can I vibecode Vernigo?

Build me a personal YouTube outlier research tool inspired by Vernigo, designed to help creators discover viral video concepts and high-performing content niches. The app should be built as a Next.js (React) single-page application using Tailwind CSS for a premium, dark-mode aesthetic with glassmorphic elements and subtle micro-animations. Use a modern color palette featuring deep midnight blues, vibrant neon accents (like electric blue or purple), and clean typography (Inter or Outfit).

**Core Concept & Data Model:**
The tool acts as a local dashboard analyzing a manually seeded list of YouTube channels. You will build a backend (Node.js/Express or Next.js API routes) connected to a PostgreSQL database.
- Models should include: `Channel` (id, name, thumbnail URL, subscriber count, assigned niche, last fetched), `Video` (id, channel_id, title, thumbnail, views, duration, publish date, outlier score), and `Folder` (for bookmarks).
- Calculate an "outlier multiplier" for each video: the video's views divided by the average views of the channel's previous 5-10 videos.

**Features & Logic:**
1. **Channel Management:** A UI to add YouTube channel IDs manually or import them via CSV. The backend will fetch the channel's recent videos using the official YouTube Data API.
2. **Dashboard & Video Grid:** The main interface should feature a highly responsive, masonry-style or grid layout displaying video cards. Each card must show the thumbnail, title, channel name, views, and the calculated outlier score highlighted with a badge. 
3. **Advanced Filtering & Sorting:** Include a sleek sidebar or slide-out panel allowing users to filter the grid by multiplier (e.g., > 3x, > 5x), total views, video duration, publish date, and specific niches. Enable sorting by highest multiplier, most viewed, or newest.
4. **Niche Analytics:** A dedicated "Niches" page that aggregates data by assigned niche. Show metrics like median views per video and overall niche momentum to help identify unsaturated topics.
5. **Collections & Bookmarking:** Users should be able to create custom folders (e.g., "Thumbnail Ideas", "High ROI Topics") and easily save videos to them using drag-and-drop or a quick save button on the video card.
6. **Automated Refresh:** Implement a background job schedule (using cron or a worker queue) to fetch the latest videos for seeded channels daily, respecting YouTube API quota limits.

**UI/UX Details & Aesthetics:**
- **Navigation:** A sleek, sticky sidebar navigation for switching between 'Dashboard', 'Channels', 'Niches', and 'Collections'.
- **Interactions:** Implement hover states on video cards that subtly scale the thumbnail, reveal action buttons (save, watch), and highlight the outlier score.
- **State Management:** Use Zustand or React Context for predictable state management across the dashboard and collections.
- **Loading States:** Use skeleton loaders for the video grid during data fetching, avoiding jarring layout shifts. Include toast notifications for success/error states (e.g., "Channel added", "API quota exceeded").

Provide a complete setup guide, including how to configure the YouTube API key, initialize the database, and start the development server. Exclude out-of-scope features like a global 10M video corpus or automated niche classification. The focus is on a stunning, functional personal dashboard.