@@ -34,7 +34,7 @@ const { content, classes, fullWidth = true }: Props = Astro.props
 const tableClass = `text-left${fullWidth ? ' w-full' : ''}`
 const captionClass = 'text-left px-5 py-3 text-content-offset'
 const headerClass = 'px-5 py-3 font-semibold'
-const theadClass = 'bg-primary text-primary-inverse'
+const theadClass = 'bg-note-offset text-primary-inverse'
 const tbodyClass = ''
 const tfootClass = ''
 ---
 
@@ -16,7 +16,7 @@ featured: true
 
 Service catalogs follow a depressingly predictable arc. At launch, you've got 90% coverage and 95% accuracy — teams are entering their services because the initiative has leadership attention. By year two, coverage has dropped to 45%, accuracy to 30%, and the catalog has become a punchline in onboarding jokes. Engineers ask in Slack instead of checking the catalog because they've learned they can't trust it.
 
-Here's what makes this worse: a catalog with 80% accurate data is more dangerous than no catalog at all. It gives you false confidence. You page the listed owner at 3 AM, confident you've got the right team, and waste twenty minutes before discovering they handed off the service six months ago. Every minute spent paging the wrong team is a minute your users are affected.
+<Highlighter>Here's what makes this worse: a catalog with 80% accurate data is more dangerous than no catalog at all. It gives you false confidence.</Highlighter> You page the listed owner at 3 AM, confident you've got the right team, and waste twenty minutes before discovering they handed off the service six months ago. Every minute spent paging the wrong team is a minute your users are affected.
 
 The fix isn't discipline or better training — it's ownership modeling that captures how teams actually work, combined with enforcement automation that keeps data accurate without relying on anyone remembering to update it.
 
@@ -31,8 +31,9 @@ But "ownership" is deceptively simple. A service might have a development team t
 The ownership model that works distinguishes between different types of responsibility: a __primary owner__ who's responsible for the service's existence and development, plus __role-specific contacts__ for specialized functions.
 
 <Table
-  variant="vertical-column-delineation-table"
+  variant="numbered-review-table"
   content={{
+    figure: 'Ownership roles and their triggers.',
     thead: {
       th: ['Role', 'Responsibility', 'When contacted'],
     },
@@ -60,7 +61,6 @@ The ownership model that works distinguishes between different types of responsi
         },
       ],
     },
-    figure: 'Ownership roles and their triggers.',
   }}
 />
 
@@ -189,8 +189,13 @@ Code: GitHub Action for catalog validation on PRs.
 For tier-1 and tier-2 services, make validation failures blocking. For tier-3 and tier-4, warn but allow the PR to merge — you want to reduce friction for less critical services while maintaining strict standards for critical ones.
 
 <Table
-  variant="vertical-column-delineation-table"
+  variant="grid-and-accent-header-table"
+  fullWidth={false}
+  classes={{
+    tbody: '!text-content text-center',
+  }}
   content={{
+    figure: 'Validation enforcement by tier',
     thead: {
       th: ['Service tier', 'Missing owner', 'Missing runbook', 'Missing dependencies'],
     },
@@ -210,11 +215,10 @@ For tier-1 and tier-2 services, make validation failures blocking. For tier-3 an
         },
         {
           th: 'tier-4',
-          td: ['Warn', '—', '—'],
+          td: ['Warn', '---', '---'],
         },
       ],
     },
-    figure: 'Validation enforcement by tier.',
   }}
 />
 
@@ -256,7 +260,11 @@ Drift detection requires API access to your identity provider, oncall system, an
 A catalog without health metrics will silently decay. Measuring catalog health requires tracking three dimensions: coverage (what percentage is cataloged), accuracy (does the data reflect reality), and freshness (when was it last updated).
 
 <Table
-  variant="vertical-column-delineation-table"
+  variant="grid-and-accent-header-table"
+  fullWidth={false}
+  classes={{
+    tbody: '!text-content',
+  }}
   content={{
     thead: {
       th: ['Metric', 'Formula', 'Target'],