page-by-pagelisted

# Page by Page Run the same review question against every page of a running application, one subagent per page, and merge what comes back into one report. The question is the parameter. The machinery — enumerate the pages, fix a shared yardstick, fan out, consolidate by cause rather than by page — does not change with it. ## Usage ``` page-by-page [app] "<question>" [--url <base>] [--scope <nav-group|prefix>] [--features] ``` ``` page-by-page myApp "how could this look cleaner? sleeker? more consistent across the ui?" page-by-page appXyz "review and improve the accessibility and mobile ux" page-by-page . "where do empty, loading, and error states fall short?" ``` - Zero args → infer the app from the current repository and the URL from its dev configuration. - No question → default to `how could this look cleaner? sleeker? more consistent across the UI?` - `--scope` → limit to one navigation group or route prefix. Use it on large apps. - `--features` → also file remediation features (Phase 7). Default is report only. For a design-specific pass that reviews **every tab within every page** as its own surface, use `frontend-design-sweep` instead. This skill's unit of work is the page. ## Applicability Applies to any project that serves a browsable interface, whatever the stack. The two genuine preconditions are a **reachable running URL** and **more than one page** to compare. A CLI, a library, or a single-screen tool has nothing for this skill to do — say so and stop.