Link Wash

Strip tracking junk from any link, instantly and entirely on your device.

Right-click any link to copy a clean version, or paste a link into the popup to see exactly what gets removed. No accounts, no analytics, no network requests.

Link Wash does one thing well: it removes tracking parameters (utm_source, fbclid, gclid, and friends) from URLs without breaking the link.

Source on GitHub · MIT licensed · zero dependencies

See it in action: paste a link and watch the trackers fall away.

Light	Dark

---

Why

Marketers append tracking parameters to links so they can follow who clicked, which campaign worked, and who shared a link. Link Wash removes that cruft, leaving a shorter, tidy, private URL, without stripping parameters the page actually needs (like id=123, v=, q=, page=).

Core principles:

• One job, done well. Clean URLs. No QR codes, no encoders, no redirect followers, no "toolkit." If a feature would need a network request, it doesn't belong here.
• 100% local & private. Zero network calls, zero analytics, zero accounts. Verifiable as zero requests in DevTools.
• Fewest permissions possible. Every permission is justified below.
• Never break a link. When unsure whether a param is tracking or functional, Link Wash keeps it.

---

Install (Load unpacked, no build step)

Link Wash is plain ES modules and CSS. There is nothing to compile.

1. Clone or download this repo.
2. Open chrome://extensions (or edge://extensions).
3. Turn on Developer mode (top right).
4. Click Load unpacked and select the src/ folder.
5. Pin Link Wash to your toolbar. Done.

To package for the store: npm run zip produces link-wash.zip.

---

How to use it

Right-click (the star feature). Right-click any link → Copy clean link. Right-click empty page space → Copy clean link to page. A small ✓ badge confirms the copy.

Popup (paste-to-clean). Click the toolbar icon, paste a link. It cleans as you type and shows:

• a summary up top (e.g. "3 trackers removed · 46% shorter") so the win registers before the details;
• a before / after comparison with removed parts struck through;
• a restore list: every removed parameter with a plain-language explanation and a toggle to keep it. Flip one on and it returns to the link, with the strike-through fading out live. Great for the odd utm or affiliate ref you actually rely on.

Use current tab fills in the page you're on. Auto-copy (below) fires only on a deliberate paste / Use current tab, never mid-typing, so it won't quietly clobber your clipboard while you edit.

Settings (in the popup's ⚙ drawer, or the full options tab):

Setting	Default	What it does
Auto-copy on clean	On	Copies the cleaned link when you paste one (or use Use current tab). Never fires while you're typing.
Aggressive mode	Off	Also strips ref-style params (see the tradeoff below).
Unwrap redirect links	Off	Pulls the real destination out of a wrapper link. Decodes the text only, never visits the link.
Your own parameters	n/a	Add any extra param name you want removed.

---

Permissions (and why each one is needed)

Link Wash requests the minimum set. It deliberately requests no host access, no tabs history, no webRequest, no cookies, and nothing network-related.

Permission	Why it's needed
contextMenus	The right-click Copy clean link items.
clipboardWrite	To write the cleaned URL to your clipboard.
storage	To save your settings and custom parameter list (chrome.storage.local).
activeTab	To read the current tab's URL on demand when you click Use current tab or Copy clean link to page. Grants access only to the active tab, only when you act.
offscreen	MV3 service workers have no DOM. A hidden offscreen document is the documented, permission-light way to copy to the clipboard from the right-click handler.

Not requested (on purpose): <all_urls>, tabs, webRequest, cookies, history, or any network/host permission.

---

Privacy statement

Link Wash makes zero network requests. Everything is computed locally.

All cleaning happens in a pure JavaScript function (src/lib/cleaner.js) that runs on your machine. There are no analytics, no telemetry, no accounts, and no remote rule updates. You can verify this yourself: open DevTools → Network while using the extension and watch it stay empty. The entire codebase is small and dependency-free precisely so you can audit it in one sitting.

Full policy: PRIVACY.md.

---

What it intentionally does NOT do (and why)

These are different products. Bolting them on would dilute the one job, add permissions, or (worst of all) require network access. Left out on purpose:

• No QR-code generation. A separate concern; adds UI noise.
• No URL encode/decode tools. That's a developer utility, not link hygiene.
• No domain/WHOIS inspector. Would need network access, a non-starter.
• No redirect checker that pings servers. Link Wash unwraps redirects by decoding the string locally; it will never visit a link. A checker that follows redirects over the network breaks the privacy promise.
• No batch-file uploads / bulk processing. Keeps the UI to one calm action.

Keeping the scope this tight is the feature.

---

Add your own rules

The denylist lives in src/lib/rules.js and is structured so adding a parameter is a one-line change:

// In TRACKING_PARAMS: [param name, plain-language explanation]
['my_tracker', 'What this parameter tracks'],

• TRACKING_PARAMS: exact param names (matched case-insensitively).
• TRACKING_PREFIXES: strip anything starting with a prefix (e.g. utm_).
• AGGRESSIVE_PARAMS: ref-style params, only stripped in aggressive mode.
• FRAGMENT_TRACKERS: known trackers inside a #a=b fragment.

Prefer not to edit code? Use Settings → Your own parameters to remove.

The ref tradeoff

Some sites genuinely route on ref/source-style params (referral and affiliate programs, share attribution that changes behavior). Because removing them can change where a link goes, Link Wash keeps them by default and only strips them when you turn on Aggressive mode. Correctness beats aggressiveness.

---

Develop & test

The cleaning engine is a pure module, so it's fully unit-tested with no framework, just Node's built-in test runner:

npm test     # runs tests/cleaner.test.mjs via `node --test`
npm run lint # zero-dependency syntax check of every JS file
npm run zip  # package src/ into link-wash.zip

There are no runtime dependencies and no build step, so the shipped code is exactly the code you read.

Project structure

src/
  manifest.json
  background/service-worker.js   context menus + clipboard + badge
  lib/cleaner.js                 pure engine: cleanUrl()
  lib/rules.js                   the denylist + explanations (extend here)
  lib/settings.js                shared settings (chrome.storage.local)
  lib/clipboard.js               MV3-safe copy via offscreen document
  lib/browser-shim.js            chrome.* namespace (Firefox port seam)
  offscreen/                     hidden clipboard helper page
  popup/                         paste-to-clean UI
  options/                       full settings tab
  icons/                         16 / 32 / 48 / 128
tests/cleaner.test.mjs
scripts/                         lint.mjs, zip.mjs (dependency-free)

---

Manual test checklist

Mirrors the automated suite plus the things only a human can click. See TESTING.md.

---

Roadmap

• Firefox port. The chrome.* namespace is funnelled through src/lib/browser-shim.js so this is a small change.
• Optional richer per-host rules (kept conservative).
• Localized parameter explanations.

That's it. No "v2 toolkit." The goal is to stay small, sharp, and finished.

---

License

MIT.