cupertino-reviewlisted

Run the fixed pipeline, in order, on the given project or scope: ``` cupertino-backwards -> cupertino-focus -> [cupertino-longevity & cupertino-integrate, side by side] -> cupertino-council -> cupertino-prototype -> cupertino-elevate -> cupertino-unbox -> cupertino-reveal ``` ## Before you start ```bash python3 "${CLAUDE_PLUGIN_ROOT}/scripts/state.py" init python3 "${CLAUDE_PLUGIN_ROOT}/scripts/state.py" set review-pipeline-active for f in backwards-done focus-output longevity-output integrate-output council-output prototype-output elevate-output unbox-output; do python3 "${CLAUDE_PLUGIN_ROOT}/scripts/state.py" clear "$f" done ``` Clearing every stage's flag before starting is deliberate: a stale `backwards-done` (or any `*-output`) left over from a **previous, unrelated** review would otherwise let this run skip a stage it hasn't actually done yet for this scope, or thread a prior scope's content into this one. Every full pipeline run starts from a clean slate — `cupertino-backwards` runs fresh every time, never reused across separate `cupertino-review` invocations even on a similar-sounding scope. Setting `review-pipeline-active` is what makes the PreToolUse hook refuse an automatic dispatch of `cupertino-cannibalize` for the duration of this pipeline — that refusal is intentional, not a bug to route around. Clear every flag when you finish (success, early stop, or the user cancels) — leaving them set would let a future run wrongly treat this one's stages as alread