A household privacy control plane, built strictly to its 2026-10-27 blueprint. One app plus a household dashboard operates five machines — Shield, Home, Removal Desk, Aliases, and Wizards — across the nine ways a household is tracked: cross-site and cross-app profiling, household and cross-device graphs, the data-broker economy, location brokering, smart-TV recognition, voice-assistant data, receipt and inbox scanning, purchase-history targeting, and hashed customer-list matching. For every practice, every person, and every device it shows one of three states in plain words — Closed, Action needed, or Can't be reached — and it keeps a statement, not a score: Seen, Closed, and Unreachable, every entry traceable to an observation. One admin runs the plan for the household; each person keeps their own removal file, because deletion rights are per-person by law, and shared devices — TVs, the router, speakers — belong to the household. Household Privacy Control is the descriptive name this first version ships under; no settled, cleared brand name exists for it yet.
How it works
- Set up the household: its name, its admin, and its state, then add its people, its devices, and the accounts and services it uses — everything stays in this browser's local storage; nothing is uploaded
- Read the Practice Map: all nine practices with their blueprint grades shown as graded, whether each applies to this household and why, its truth label, the step-by-step control workflow of real settings and real pages on companies' own sites, and a permanent box naming what can't be reached
- Work the wizards: TV recognition off per brand, voice history reviewed and deleted with auto-delete set, inbox grants audited grant by grant on Google's and Microsoft's own permissions pages, location permissions audited, retailer ad preferences set
- Run the Removal Desk per person: sign the permission artefact once, find the listings before removing anything, prepare each broker's deletion request — Californians file the state's free DROP request first — record the honest status, and re-check on a cycle
- Use aliases from today: a unique alias email for every new signup, so future hashed-list matching has nothing real to match, and a leaked alias names the service that leaked it
- Read Your Statement: Seen, Closed, and Can't be reached, with its counts reconciled in the open — and no score, ever
Nine practices, graded honestly — no score, ever
The blueprint graded all nine practices before anything was built, and the app shows those grades exactly as graded, character for character. No practice grades fully findable and controllable — none. Seven grade Partial and two grade Prevention-only, and the honest product story is strong prevention plus partial remediation plus verified guidance, never “find and delete your surveillance.”
There is no privacy or exposure score anywhere in this product — no number with no honest denominator, computed from checks chosen because they move. Instead the control plane keeps the three evidence buckets: Seen (things directly observed, each with its evidence attached), Closed (verified state changes, each with how it was verified), and Unreachable (named in plain language, permanently visible, never averaged away). Server-to-server sharing, the household graph on other companies' servers, first-party purchase profiles, and data already sold sit in that last bucket by architecture, not by roadmap.
- Cross-site & cross-app profiling — Partial
- Household & cross-device graphs — Prevention-only
- The data-broker economy — Partial
- Location brokering — Partial
- Smart-TV recognition (ACR) — Partial
- Voice-assistant data — Partial
- Receipt & inbox scanning — Partial
- Purchase-history targeting — Partial (weak)
- Hashed customer-list matching — Prevention-only
The five machines under the control plane
No single mechanism reaches more than a few of the nine practices, so the product is a control plane operating five machines, each with an honest truth label it is allowed to claim. The control plane's job is state, not magic — and the control plane never sells, shares, or monetises the data it touches; the business model is subscription for software and labour, never data.
- Shield — on-device blocking and signals: tracker blocking, a managed encrypted-DNS profile, and Global Privacy Control guidance. It uses the phone's single VPN slot, so it cannot run alongside your own VPN. Its truth label: blocks the known, visible third-party layer on this device — never “stops tracking.”
- Home — the household network layer: a managed resolver configuration at the router and per device, with per-device visibility. Its log is a floor, not a total — devices with their own encrypted DNS bypass it, and the product says so per device when it sees a device go dark.
- Removal Desk — the rights service: a signed permission once, then listing discovery with proof, broker-by-broker deletion and opt-out requests, Californians handed to the state's free DROP system first (never charged for what DROP does free), and re-checks on a cycle because re-listing is routine. It reports this household's own measured removal rate against the Consumer Reports benchmark — paid removal measured about 35% of found listings removed in four months, the best service about 68%, careful do-it-yourself about 70% — never a coverage count.
- Aliases — prevention: a unique alias email per service over existing providers, with leak attribution when an alias starts receiving spam. Its truth label: protects the next signup, not the last ten years.
- Wizards — guided settings that end in a verified state, not a “good luck” article: TV recognition off per brand, confirmed by the TV's recognition traffic going quiet in the Home log; voice-history review, deletion, and auto-delete setup; the Google and Microsoft inbox-grant audit, walked grant by grant; location-permission audits; retailer ad-preference walkthroughs. Voice-data deletion and the inbox-grant audit are the two capabilities the blueprint found no product in the field offering at all.
Where it stands — the honest state
Household Privacy Control is built and verified to its blueprint: a single-file, test-mode web app with no dependencies, all state in the browser's local storage. Its verification harness exercises the shipped app end to end — 58 of 58 checks pass with zero script errors, including the grade-fidelity check: all nine blueprint grades read back from the running app character for character, 9 of 9. State honesty is enforced by construction and tested: a Prevention-only practice can never display Closed, even with every prevention step done.
Test mode is a discipline, not a label: nothing is submitted, no account is connected, and every request is prepared for the user to send themselves. The broker-letter flow asks for the broker name through a plain prompt in this version — functional, and plainly a first-version seam. The Android Shield appears in this version as its honest guidance layer only — advertising-ID deletion, the tracking consent gate and App Privacy Report, and Global Privacy Control browser guidance, with the one-VPN-slot rule stated plainly. A web page does not block traffic, and this one does not pretend to.
The rendered layout has been checked in a real browser at phone and desktop widths (390px and 1280px): no horizontal overflow on the main views, the full workflow walk completed in that browser with no console errors, and the captures were inspected. Earlier verification was code-only — the build machine's managed browser is blocked by policy from launching — until a bundled-browser pass closed that gap in the same cycle that built this page. A look on the founder's own phone remains his device step, as it is for every product here.
The native platform builds are later blueprint steps, not part of this version: the Android Shield VPN build and the iOS device-wide firewall path come after, in the blueprint's own sequence. And the name is open: Household Privacy Control is a descriptive first-version name, not a settled or cleared brand name — the blueprint's naming lane was left open, and this page claims no cleared name.
Pricing has not been decided. The blueprint carries pricing figures as design, labeled as design in the blueprint itself; they are not a ruling, and no price is quoted on this page. There is nothing to buy here; today this door does one honest thing: it takes an interest list.
What a first real use still needs
Three steps stand between this build and a first real-world use, named plainly:
First, the founder's go before any real household data beyond test entries is used, and before any request is actually sent to a broker. This build prepares; sending is the user's own act, outside test mode.
Second, the native platform builds the blueprint sequences later — the Android Shield VPN build (with its Play VpnService declaration) and the iOS firewall path (organisation enrolment) — plus wizard-library maintenance as TV firmware and app menus move, because the wizards are a living library, not a one-time build.
Third, a settled product name. The blueprint's naming lane was left open: this version ships under the descriptive name Household Privacy Control and claims no cleared brand name.
Join the Household Privacy Control interest list
This is an interest list — interest only. Leave your name and email, and we will write to this list when there is news about Household Privacy Control. Joining does not create an account, and it is not a purchase — pricing has not been decided, and nothing is for sale on this page.