A personal knowledge capture tool that turns any webpage into structured, searchable content — saved to your own library.
Gatherly is a personal knowledge library. Paste any URL and Gatherly automatically fetches the page, strips away noise (ads, nav, clutter), and stores the content as clean Markdown — with metadata like title, author, publish date, and OG image. It then uses AI to summarize the content and tag it automatically, so your saved pages become a searchable, structured knowledge archive scoped privately to your account.
Save any webpage by providing a URL. The system fetches and processes the page content automatically.
Web pages are cleaned and converted into structured Markdown, removing ads, navigation clutter, and noise.
Automatically extracts useful metadata from each page:
- Title
- Author
- Published date
- Open Graph image
Each saved page is automatically summarized using AI, giving you a concise overview of the content without having to re-read the full article.
AI analyzes the content of each saved page and automatically assigns relevant tags, making your library self-organizing and easy to browse by topic.
Use the Discover route to fetch a curated set of links for any topic of interest. Explore new content and add it directly to your knowledge library without manually hunting for URLs.
Import multiple URLs at once and process them in parallel for faster knowledge capture.
Discover multiple pages from a website automatically — useful for exploring documentation, blogs, and research sources.
Each saved page moves through a transparent processing lifecycle:
| Status | Description |
|---|---|
processing |
Page is being fetched and parsed |
completed |
Content successfully extracted and stored |
failed |
An error occurred during capture |
Captured pages are stored as structured records in the database — searchable, revisitable, and reusable.
All captured content is stored as Markdown, making it easy to render in apps, export, or reuse in docs and notes.
Each user's saved content is securely scoped to their account via Better Auth.
Built with TanStack Start server functions for type-safe, fast server-side operations.
Landing page content is rendered server-side, producing fully hydrated HTML for search engines.
| Layer | Technology |
|---|---|
| Framework | TanStack Start |
| Routing | TanStack Router (file-based) |
| Forms | TanStack Form |
| Auth | Better Auth |
| Database ORM | Prisma |
| Database | Neon PostgreSQL |
| Web Scraping | Firecrawl API |
| AI (Summarization & Tagging) | Claude / OpenAI |
| Validation | Zod |
| UI Components | shadcn/ui |
| Styling | Tailwind CSS v4 |
| Notifications | Sonner |
| Analytics | Vercel Analytics |
| Language | TypeScript |
- Node.js 18+
- pnpm
git clone https://github.com/Srinivaskoruprolu007/Gatherly.git
cd Gatherly
pnpm installCreate a .env file in the root:
DATABASE_URL=your_neon_postgresql_connection_string
BETTER_AUTH_SECRET=your_auth_secret
BETTER_AUTH_URL=http://localhost:3000
FIRECRAWL_API_KEY=your_firecrawl_api_key
# AI provider (whichever you use)
ANTHROPIC_API_KEY=your_anthropic_api_key
# or
OPENAI_API_KEY=your_openai_api_keypnpm devpnpm build
pnpm start| Command | Description |
|---|---|
pnpm dev |
Start development server |
pnpm build |
Build for production |
pnpm start |
Start production server |
pnpm test |
Run tests with Vitest |
pnpm lint |
Lint with ESLint |
pnpm format |
Format with Prettier |
pnpm check |
Run lint + format check together |
src/
├── routes/ # File-based routes (TanStack Router)
│ ├── __root.tsx # Root layout — head, theme, toaster, analytics
│ ├── _authed/ # Authenticated route group
│ │ ├── discover/ # Topic-based link discovery
│ │ └── ...
│ └── ...
├── components/ # Shared UI components (shadcn/ui)
├── context/ # React context providers (ThemeContext)
├── lib/ # Utilities, constants, server functions
└── styles.css # Global styles (Tailwind)
prisma/
└── schema.prisma # Database schema (Neon PostgreSQL)
Gatherly separates concerns into distinct layers:
URL Input
│
▼
Firecrawl API ← fetch, clean, extract metadata
│
▼
AI Pipeline ← summarize content, auto-generate tags
│
▼
Prisma / Neon ← store structured records
│
▼
UI (TanStack Router) ← render, search, discover
This separation allows the platform to grow into a full knowledge management system — with search, filtering by tag, export, and more.
Deployed on Vercel with automatic deployments on every push to main.
🔗 https://gatherly-bice.vercel.app
MIT