Building•By John Iseghohi•Aug 5, 2026•5 min read

How to Build a Chrome Extension in a Weekend (2026)

Ship a Manifest V3 Chrome extension in one weekend with AI tools — idea pick, build steps, Web Store publish, and how indies actually monetize in 2026.

Who This Is For

You want a product you can ship without spinning up a full SaaS backend.

You've heard Chrome extensions are "easy money" and want the honest weekend path.

You're comfortable pasting prompts into Cursor or Claude Code, but you've never touched manifest.json.

If that's you, keep reading.

Why Extensions Still Work in 2026

The Chrome Web Store still has well over 100,000 published extensions. Roughly 60% are built by solo indie developers. The catch: the median extension has about 18 users, and ~70% never clear 100 installs.

So the opportunity isn't "publish anything." It's ship a narrow, painful, browser-native fix — preferably where an older Manifest V2 tool died and left angry 1-star reviews behind.

Good news on cost: Google still charges a one-time $5 developer registration fee. Not $99/year. One coffee for unlimited future listings under that account (up to the account limit).

Need an idea first? Browse curated problems on startup ideas and ask: "Does this pain show up inside the browser?"

The One-Feature Rule

Your weekend dies if you try to build a "platform."

Pick one action the user can complete in under 10 seconds:

• Save the current page title + URL to a list
• Summarize the open tab
• Highlight and export selected text
• Check SEO meta tags on the current page
• Block one distracting site during focus hours

If you can't say the job in one sentence, cut scope.

Tools You Need

• Cursor or Claude Code — generate Manifest V3 scaffolding and fix service-worker gotchas
• Chrome — load unpacked for local testing
• GitHub — optional, but useful once you iterate
• $5 — Chrome Web Store developer fee when you're ready to publish
• Stripe / Lemon Squeezy / ExtensionPay — only if you add paid unlock later (Chrome Web Store's old in-store payments are gone)

You do not need React on day one. Plain HTML/CSS/JS is enough for most MVPs.

Weekend Framework

Friday Night (45–60 min): Idea + Permissions Lock

1. Search the Chrome Web Store for your idea.
2. Read the 1-star reviews on the top 3 competitors. That's your feature list.
3. Write a sticky-note brief:

"When I'm on [site/context], I want to [action], so I can [outcome]."

4. List the minimum permissions. Prefer activeTab + storage. Avoid tabs and broad host access (*://*/*) unless you truly need them — broad permissions slow review and scare installs.

Saturday (6–8 hours): Build the Engine

Ask your AI coding tool for a Manifest V3 project with:

• manifest.json (manifest_version: 3)
• Popup UI (one button, one status line)
• Service worker for background logic
• Content script only if you must touch the page DOM
• chrome.storage.local for persistence (never put state in a global variable — service workers die when idle)

MV3 traps AI still gets wrong:

1. Service workers are ephemeral — use chrome.storage.session or chrome.storage.local, not in-memory maps
2. Async message listeners must return true to keep the channel open
3. No remote code execution — keep scripts in the package
4. Dynamic SPAs (GitHub, Gmail, Notion) need a MutationObserver or you'll inject the UI once and lose it on client-side navigation

Test loop:

1. Open chrome://extensions
2. Enable Developer mode
3. Load unpacked → select your folder
4. Click the toolbar icon
5. Break it on purpose (offline, weird page, missing API key)

By Saturday night you should have a working core action, even if the UI is ugly.

Sunday Morning: Polish + Store Assets

• Icons at 16, 48, and 128px
• Short description under 132 characters
• 2–3 screenshots at 1280×800 that show the outcome, not the settings page
• Privacy policy URL (required if you send any data off-device — even to an LLM API)

Sunday Afternoon: Submit

1. Zip the build so manifest.json is at the root of the zip (most common rejection)
2. Pay the $5 if you haven't
3. Upload → fill listing → submit
4. Simple extensions with minimal permissions often clear in 1–3 business days

While you wait, post a demo GIF in one niche community. Don't spam five subs.

Monetization Reality Check

Chrome Web Store payments are deprecated. Indies charge via external processors.

Directional market mix for monetized extensions: freemium/subscription ~35%, ads ~25%, one-time ~15%. About 30% of monetized extensions earn enough to cover costs; the long tail earns nothing.

Practical weekend default:

1. Ship free
2. Watch weekly retention for 2–4 weeks
3. Gate a repeated workflow behind $5–15/mo (or a cheap one-time unlock)

If nobody returns after day 3, a paywall won't save you.

What to Build First (5 Ideas)

1. SEO meta checker — title, description, H1s, OG tags in one click
2. Tab session saver — save/restore named tab groups
3. Highlight → note — selection + source URL to a sidebar
4. Price drop watcher — one retailer, one alert channel
5. Focus site blocker — one allowlist, one schedule

Each is a weekend. Portfolio beats perfection.

Stuck picking? Start from a validated problem on startup ideas and force the "browser surface" filter.

FAQ

Do I need to know JavaScript?

You need to read errors and describe intent clearly. AI writes most of the boilerplate. You own testing and permission hygiene.

Is Manifest V2 still an option?

No. New launches are Manifest V3. Treat MV2 as archaeology — and as a hunting ground for abandoned extensions with leftover demand.

Can I make money with under 1,000 users?

Yes, if the job is high-frequency and B2B/professional. A few hundred retained users at $8/mo beats 50,000 tire-kickers.

TL;DR

Lock one browser job. Build Manifest V3 with minimal permissions. Load unpacked Saturday, polish Sunday, pay $5, submit. Monetize only after retention shows up. Extensions are still one of the fastest weekend surfaces left — if you stay ruthless on scope.

When you're ready for the next product after the extension wedge, grab a fuller SaaS-shaped idea from startup ideas and reuse the same ship-in-a-weekend muscle.