What we check, and what currently fails.

Every app here claims to run, to keep your data on your device, and to download as something you can build. Those are checkable claims, so we check them, and this page publishes the results — including the ones that fail. A catalog that only shows its passes is asking you to take its word for it.

40 apps checked

Last run August 2, 2026, against commit c11c184. Checks are re-run before release and expire after 30 days — stale evidence stops the build rather than quietly ageing on the page.

None failing

Every app passes every check it has been put through. When one does not, it says so here and on its own page, in red.

Nothing unclaimed

Every check in the standard has been run against every app. Nothing on a listing is asserted without evidence behind it.

The checks

Demo renders
A headless browser loaded the demo, waited for it to render, and captured the screenshot shown on its page. The picture is evidence, not artwork.
Package installs and builds
The published zip was extracted into an empty directory, installed from scratch, typechecked, and built. This is the claim the whole catalog rests on.
Source package complete
The zip contains the files a runnable Next.js project needs — package.json, tsconfig, next config, and the app entry points.
Agent instructions included
CLAUDE.md, AGENTS.md, the Cursor rules, and the agent handbook are present inside the package, not just promised on the page.
No network calls in app source
A static scan of the app source found no fetch, XHR, WebSocket, eval, cookie access, or external URL. This is what 'your data stays in your browser' has to mean to be worth saying.
License matches the package
The license named on the listing is the license in the LICENSE file inside the package.

Every app, every check

AppDemo rendersPackage installs and buildsSource package completeAgent instructions includedNo network calls in app sourceLicense matches the package
Kanban Board Template Apppasspasspasspasspasspass
Eisenhower Matrix Task Managerpasspasspasspasspasspass
Personal Operating Systempasspasspasspasspasspass
Client Portalpasspasspasspasspasspass
Personal CRMpasspasspasspasspasspass
SaaS Foundationpasspasspasspasspasspass
Marketplace Foundationpasspasspasspasspasspass
Agent Workbenchpasspasspasspasspasspass
Invoice & Estimatespasspasspasspasspasspass
Fitness Coach Platformpasspasspasspasspasspass
Knowledge Basepasspasspasspasspasspass
Internal Admin Dashboardpasspasspasspasspasspass
Newsletter Platformpasspasspasspasspasspass
Content Calendarpasspasspasspasspasspass
Job Boardpasspasspasspasspasspass
Habit Trackerpasspasspasspasspasspass
Expense Trackerpasspasspasspasspasspass
Subscription Trackerpasspasspasspasspasspass
Workout Logpasspasspasspasspasspass
Recipe Boxpasspasspasspasspasspass
Job Application Trackerpasspasspasspasspasspass
Bug Trackerpasspasspasspasspasspass
Travel Plannerpasspasspasspasspasspass
Breathworkpasspasspasspasspasspass
Meditation Apppasspasspasspasspasspass
Garmin Watch Kitpasspasspasspasspasspass
Baby Trackerpasspasspasspasspasspass
Car Maintenance Logpasspasspasspasspasspass
Book Trackerpasspasspasspasspasspass
Debt Payoff Plannerpasspasspasspasspasspass
Just in Casepasspasspasspasspasspass
Foundation Workspacepasspasspasspasspasspass
Time Trackerpasspasspasspasspasspass
Budget Plannerpasspasspasspasspasspass
Flashcardspasspasspasspasspasspass
Freelance Rate Calculatorpasspasspasspasspasspass
Blood Pressure Logpasspasspasspasspasspass
Planting Calendarpasspasspasspasspasspass
Drawdown Monitorpasspasspasspasspasspass
Resume Builderpasspasspasspasspasspass

The same results are published as standards.json if you would rather read them as data. Hover a cell for what the check actually measured.

What these checks do not tell you

They are structural, not editorial. A passing app is one that runs, builds, keeps its data local, and ships what it says it ships. None of that makes it well designed, useful to you, or free of bugs — those are judgements you make by opening the demo, which is why every demo is the real app rather than a video of one.

Accessibility, dependency freshness, and a scored review of the agent handoff are not in this list yet. When they are, they will appear here with their results, including the failures.