## Context

The `frontend/src/turbo/` layer glues Turbo Drive and Angular together but Begin of the insertionhadEnd of the insertion Begin of the deletion hasEnd of the deletion almost no unit coverage — the only spec Begin of the insertionwasEnd of the insertion Begin of the deletion isEnd of the deletion `helpers.spec.ts` (progress bar). Every integration function (`addTurboEventListeners`, `addTurboGlobalListeners`, the `register*StreamAction` functions, the Angular wrapper) Begin of the insertionhard-boundEnd of the insertion Begin of the deletion hard-bindsEnd of the deletion to the global `document` and Begin of the insertionreturnedEnd of the insertion Begin of the deletion returnsEnd of the deletion `void`, Begin of the insertiongiving vitestEnd of the insertion Begin of the deletion so there isEnd of the deletion no Begin of the insertionwayEnd of the insertion Begin of the deletion seamEnd of the deletion to drive them Begin of the insertionor tear listeners down between cases.End of the insertion Begin of the deletion from a test.End of the deletion The only coverage of real navigation behaviour Begin of the insertionwasEnd of the insertion Begin of the deletion isEnd of the deletion one expensive Selenium spec (`spec/features/notifications/turbo_angular_navigation_spec.rb`). The most upgrade-fragile code in the frontend therefore Begin of the insertionhadEnd of the insertion Begin of the deletion hasEnd of the deletion the least coverage.

The sharpest symptom Begin of the insertionwasEnd of the insertion Begin of the deletion isEnd of the deletion `frontend/src/turbo/turbo-navigation-patch.ts`: a monkeypatch of Turbo internals (`FrameElement.delegateConstructor.prototype.proposeVisitIfNavigatedWithAction`) working around https://github.com/hotwired/turbo/issues/1300. It Begin of the insertioncarriedEnd of the insertion Begin of the deletion carriesEnd of the deletion `@ts-nocheck` and `/* eslint-disable */`, so neither tsc nor eslint Begin of the insertionwatchedEnd of the insertion Begin of the deletion watchesEnd of the deletion it, and it Begin of the insertionhadEnd of the insertion Begin of the deletion hasEnd of the deletion zero tests — yet it runs on every frame-with-action visit. The patched symbol still exists in Turbo 8.0.23, so the patch still binds and Begin of the insertionisEnd of the insertion Begin of the deletion mayEnd of the deletion still Begin of the deletion beEnd of the deletion load-bearing; it Begin of the insertioncould notEnd of the insertion Begin of the deletion cannotEnd of the deletion be safely deleted without a spec that reproduces the underlying symptom.

## Goal

Begin of the insertionAdd unitEnd of the insertion Begin of the deletion Introduce aEnd of the deletion test Begin of the insertioncoverageEnd of the insertion Begin of the deletion seamEnd of the deletion across `frontend/src/turbo/`, then use it to Begin of the insertionsettle whetherEnd of the insertion Begin of the deletion makeEnd of the deletion the navigation monkeypatch Begin of the deletion a decidable delete rather than a guess.

## Tasks

- Adopt one convention: every installer takes its event target as a parameter (e.g. `addTurboEventListeners(target = document)`) and returns a teardown handle where it owns mutable state, so the directory becomes jsdom-drivable.
- Add vitest specs that construct a target, dispatch synthetic `turbo:*` events, and assert effects (dialog close on `turbo:submit-end`, custom stream actions, nonce scrubbing) — extending the fake-timer pattern already used in `helpers.spec.ts`.
- Write a behavioural spec reproducing the symptom `turbo-navigation-patch.ts` works around (https://github.com/hotwired/turbo/issues/1300).
- With that spec in place, test whether the patchEnd of the deletion is still Begin of the insertionneeded.End of the insertion Begin of the deletion load-bearing on Turbo 8.0.23. If the spec stays green without it, delete `turbo-navigation-patch.ts`. If not, move it behind a named `applyFrameVisitPatch()` seam pinned to the Turbo version, with the spec as the regression guard and a comment recording the upstream issue state.End of the deletion

## Notes

Relates to #DREAM-514 — this jsdom test harness is the precondition for the bootstrapping/routing rework that ticket is on hold over. Landing it improves the Begin of the insertionintegration'sEnd of the insertion Begin of the deletion integration'sEnd of the deletion testability without committing to the contentious redesign.