Lock coverage-floor gate at 90% QASP target (epic top-off)

[fpigeonjr]

Parent

#576

What to build

Final top-off slice for the coverage epic: once the preceding coverage slices have landed, re-measure current coverage (post-Vitest-migration, post all spec additions) and lock the ratchet at the QASP target — 90% across statements, branches, functions, and lines — by running the coverage:bump script and committing the resulting coverage-floor.json on its own.

If 90% is not fully reachable across all four metrics once the preceding slices land (e.g. some experimental/ internals remain thin), lock the floor at whatever is achieved, explicitly note the shortfall per metric in this issue, and file a follow-up issue for the remaining gap rather than blocking indefinitely on it. The QASP hard floor is 80% new-code coverage; 90% is the target, not an absolute gate for closing this epic.

Acceptance criteria

• Coverage re-measured after all preceding coverage slices (EPIC: Raise sam-ui-elements test coverage #576's sub-issues) have merged
• coverage-floor.json bumped via coverage:bump and committed as its own standalone commit/PR
• All four metrics (statements/branches/functions/lines) at or above 90%, OR each shortfall is explicitly documented with a follow-up issue filed
• CI passes with the raised floor
• Parent epic (EPIC: Raise sam-ui-elements test coverage #576) acceptance criteria reviewed and closed out if fully met

Blocked by

• Add ratcheting coverage-floor gate to CI #627
• Add specs for autocomplete and autocomplete-multiselect form controls #628
• Add specs for date, date-range, date-time, and time form controls #629
• Add specs for upload and upload-v2 form controls #630
• Add specs for remaining under-covered form controls #631
• Add specs for ui-kit components (modal, image, data-table sort) #632
• Add specs for remaining ui-kit gaps (layout, form-templates, directives, utilities, wrappers, pipes, elements, filters) #633
• Add specs for experimental overlay internals #634
• Add specs for experimental aria abstract base classes #635
• Add specs for remaining experimental components #636

[fpigeonjr]

Re-measured coverage after all ten blocking sub-issues (#627–#636) merged into master, and locked the ratchet via npm run coverage:bump in a standalone commit.

Measured coverage (post-migration, post all spec additions)

metric	measured	floor (previous)	floor (new)	90% target
statements	88.73%	53.56%	88.73%	short by 1.27 pts
branches	78.65%	39.77%	78.65%	short by 11.35 pts
functions	85.90%	50.31%	85.90%	short by 4.10 pts
lines	88.72%	52.84%	88.72%	short by 1.28 pts

None of the four metrics reach the 90% QASP target yet. Per the QASP standard, 80% new-code coverage is the hard floor and 90% is the target, not an absolute gate for closing this epic — so the floor is locked at the current measured coverage rather than blocked indefinitely.

Branches is the clear priority for a follow-up slice (11.35-point gap vs. ~1.3 for statements/lines and 4.1 for functions).

Filed the follow-up: #669, with the specific thin-branch-coverage modules identified from this measurement (experimental/patterns/layout/components/**, banner.component.ts, header-next, tab-nav-bar, pagination, short-date pipe, international-phone, listbox, sidenav service, sort-header, multiselect-dropdown, and the largest single opportunity — experimental/hierarchical/autocomplete at 82 branches / 49%).

AC status

• Coverage re-measured after all preceding coverage slices merged
• coverage-floor.json bumped via coverage:bump, committed as its own standalone commit
• Shortfall per metric explicitly documented above; follow-up Close remaining coverage gap to 90% QASP target (branches lag furthest) #669 filed
• CI passes with the raised floor — pending PR/CI run
• Parent epic (EPIC: Raise sam-ui-elements test coverage #576) reviewed — see separate comment there