Feedblocker. Chrome extension that blocks feeds on LinkedIn and YouTube.

Context

Role: Solo maker: product + design + build + release

Platform: Chrome extension

Timeline: ~2 months

Tools / system: Figma, Cursor, ChatGPT, Gemini, HTML/CSS/JS, DevTools

Constraints: Limited programming background, first “vibecoding” project

Brief

Ship a real, installable Chrome extension that blocks distracting feeds and replaces them with a clear interruption message—while building an AI-assisted workflow to independently deliver a 0→1 product and practice engineering skills.

It's ready and it's free to install on Chrome Store 🤩

Process Approach

Converted intent into build steps with AI: iterated PRDs into small, testable slices instead of trying to “spec everything perfectly.”

Converted intent into build steps with AI: iterated PRDs into small, testable slices instead of trying to “spec everything perfectly.”

Adopted an incremental engineering loop (after an early failure with PRD): atomize → generate/edit → read code → test in DevTools → repeat.

Adopted an incremental engineering loop (after an early failure with PRD): atomize → generate/edit → read code → test in DevTools → repeat.

Built dynamic-page resilience: combined URL rules + DOM selectors + observers to keep blocking active as pages and tabs change.

Built dynamic-page resilience: combined URL rules + DOM selectors + observers to keep blocking active as pages and tabs change.

Shipped v1, then planning to bug-first hardening to improve stability and expand to support more social media.

Shipped v1, then planning to bug-first hardening to improve stability and expand to support more social media.

Outputs

Figma UI concept for the browser extension.

Blocked Linkedin newsfeed by a production version of the extension.

Extension codebase JS samples.

Outcomes

0→1 shipped: extension published and installable from the Chrome Web Store. Use it daily.

Practiced Vibe Coding Approach: implemented a working extension with URL-aware targeting, DOM selection, MutationObserver-style behavior, and persistent state.

Clear next step identified: reliability fix for persistence/state edge cases, then scale the same approach to Facebook / Instagram / TikTok.

“Worked with Gleb on a browser extension — he turned a “hmm, maybe this?” idea into a clean, shippable product in no time. Sharp mix of UX and product thinking, and genuinely fun to work with.”