docs/CI.md's workflow summary table (line 11) lists the Lighthouse workflow's "Blocking?" column as Warn. In reality, lighthouserc.json's assert.assertions configures categories:accessibility, categories:best-practices, and categories:seo as "error" (which Lighthouse CI treats as build-failing) and only categories:performance plus the Core Web Vitals metrics as "warn" — matching docs/performance/LIGHTHOUSE.md's own table one directory over, which correctly states "Accessibility | ≥ 90 | error (blocks merge)", "Best Practices | ... | error (blocks merge)", "SEO | ... | error (blocks merge)". A contributor reading only docs/CI.md's summary table would incorrectly conclude a Lighthouse accessibility regression can't block their PR, when three of its four categories actually can. Acceptance criteria: update docs/CI.md's table to reflect the real per-category blocking behavior (e.g. "Mixed — a11y/best-practices/SEO block, performance warns") instead of a single blanket "Warn".
docs/CI.md's workflow summary table (line 11) lists the Lighthouse workflow's "Blocking?" column as
Warn. In reality,lighthouserc.json'sassert.assertionsconfigurescategories:accessibility,categories:best-practices, andcategories:seoas"error"(which Lighthouse CI treats as build-failing) and onlycategories:performanceplus the Core Web Vitals metrics as"warn"— matching docs/performance/LIGHTHOUSE.md's own table one directory over, which correctly states "Accessibility | ≥ 90 | error (blocks merge)", "Best Practices | ... | error (blocks merge)", "SEO | ... | error (blocks merge)". A contributor reading only docs/CI.md's summary table would incorrectly conclude a Lighthouse accessibility regression can't block their PR, when three of its four categories actually can. Acceptance criteria: update docs/CI.md's table to reflect the real per-category blocking behavior (e.g. "Mixed — a11y/best-practices/SEO block, performance warns") instead of a single blanket "Warn".