What Discovery is

Discovery is a personal curation site - a public-readable second brain that stores the tools, hacks, libraries, threads, and resources I find worth keeping. It exists to solve two problems at once: fast personal lookup, and a quietly indexed corner of the web that other people can discover when they search for the same things.

How it works

Behind the scenes, an automated content ingestion pipeline scans Reddit, GitHub, YouTube, and a hand-picked list of blogs. When something worth keeping shows up, it gets appended to a flat JSON file in this repo. A push triggers a Vercel deploy. There is no database, no admin UI, no CMS. Each entry can optionally have a longer MDX writeup, but the JSON is always the source of truth.

The shape of an entry

Every entry has a title, a description, a source URL, a category (tool, hack, library, video, thread, article), one or more tags, and an optional personal note explaining why I saved it. That last field is the point of the whole site - anyone can bookmark a link, but the note is what turns a bookmark into something worth coming back to.

The numbers, today

• 78 entries saved
• 181 tags in active use
• Built with Next.js, statically exported, hosted on Vercel - zero runtime cost

Where to start

The homepage shows the latest and a featured pick. Browse gives you full search and filters. Or drop into a tag like #mcp or hacks to wander.

Built and maintained by Nirav Joshi. Source code lives on GitHub.