Research / updated September 4, 2026

How we test Cloak.
And what passes.

Repeatable tests / clear limits

Proof should be rerunnable.

Each fixture starts with a tracked shopping URL or browser-visible request, runs Cloak's packaged logic, and checks both the protection result and the shopping state that must survive. This is a controlled baseline—not a promise that every live store will stay unchanged.

28shopping, search, product, cart, and checkout URL fixtures
66named blocklist rules documented in the glossary
117named request domains represented by those rules
01

Clean the visible URL

Remove supported click IDs, referral paths, recommendation baggage, and campaign tags while keeping the product or search usable.

02

Keep shopping intent intact

Verify that useful search terms, filters, variants, quantities, carts, currency, and checkout state survive.

03

Count actions honestly

Keep trackers, ads, URL tags, browser-data clears, and selected recognition reductions in separate categories.

04

Reject only clear optional consent

Verify that consent assistance prefers recognizable reject, necessary-only, and opt-out controls without selecting accept-all or ambiguous choices.

05

Reduce selected recognition reads

Probe covered graphics, audio, fonts, media devices, hardware, language, battery, and browser hints while preserving page-safe, stable behavior.

06

Keep Cloak Check local

Verify clean offer identity, visible-price capture, local timelines, search and sorting, plus an alert when the current price exceeds a lower local observation.

07

Exercise the access lifecycle

Check trial timing, prompt cooldowns, persistent dismissal, paid suppression, opaque entitlement verification, and paused protection after access expires.

Marketplaces

Amazon, Walmart, eBay, AliExpress, Alibaba, and Temu.

Retail product pages

Target, Best Buy, Home Depot, Lowe's, Costco, Ace Hardware, and Fashionphile.

Commerce platforms

Shopify-style, WooCommerce, Salesforce Commerce Cloud, Adobe Commerce, and BigCommerce flows.

Search and discovery

Search intent stays while supported click tracking is removed.

What this does not prove.

The benchmark does not prove lower prices, price discrimination, total tracker blocking, server-side deletion, or permanent compatibility. It shows what the tested build did against named fixtures on the stated date.