Make every path accessible.
The post says this testing has already taken two days. Treat it as the first finish-line gate before the platform surfaces that follow.
The app can be basically done and still need the surfaces that make it feel native, discoverable, accessible, and supportable.
VoiceOver / rotor, Spotlight, updates, AppleScript, help pages, and preferences are the remaining areas named in the supplied post.
Check off the workstreams as they are actually completed in the product. This tracker records status; it does not simulate platform testing.
The post says this testing has already taken two days. Treat it as the first finish-line gate before the platform surfaces that follow.
Integration is named as the next workstream after accessibility.
Update checking is part of the native product expectation still to be addressed.
Support for scripting is explicitly in the remaining surface area.
Support content is part of making the app feel finished for real users.
The source ends while naming preferences, so the exact preference scope remains intentionally open.
The source names VoiceOver and rotor, Spotlight, app updates, AppleScript, help pages and preferences, six separate checkbox surfaces. Three checked out of six gives fifty percent. Bars share five hundred forty pixels for the full checklist, so three checked uses two hundred seventy pixels. Checking a surface is a manual status assertion; this page does not run accessibility tests, native integrations, update checks or scripting tests. Six independent binary checkboxes have sixty four combinations, while the visible checked count can only be zero through six, seven totals. Bars compare sixty four combinations with seven possible counts at eight pixels per item. Distinct combinations with the same count have the same meter width. The meter does not weight accessibility more heavily, assess implementation depth or confirm that any workstream was actually finished. The supplied post leaves the exact preference scope open. The actual text handoff exports one row for each of the six named surfaces, marking checked with x and unchecked with a blank. Bars compare six total rows with three marked rows for a hypothetical halfway checklist at eighty pixels per row. The full local object stores all six true or false values and reload restores those choices. There is no Reset control in the source. The final export boundary says the tracker claims no platform integration or accessibility test was performed.