sync: SEO canonical URLs, curation improvements, batch scripts, stats fix

- Dynamic canonical URLs and alternates for all pages (SEO fix)
- Shared seo.ts utility for canonical path generation
- Improved duplicate detection: standalone > aggregator priority, forks/created_at tie-breakers
- browseReadyFilter now includes unique aggregator skills
- Stats aligned with browseReadyFilter (was showing 16k instead of 62k+)
- Re-review mechanism: content_hash changes trigger needs-re-review
- review_status field added to skills schema
- Batch security scan and quality score scripts
- Indexer Dockerfile updated for curation deps
- Removed roadmap from README

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

packages/db/src/schema.ts

@@ -79,6 +79,7 @@ export const skills = pgTable(
     isDuplicate: boolean('is_duplicate').default(false),
     canonicalSkillId: text('canonical_skill_id'), // points to the "original" if this is a duplicate
     repoSkillCount: integer('repo_skill_count'), // cached count of skills in same repo
+    reviewStatus: text('review_status').default('unreviewed'), // unreviewed → auto-scored → ai-reviewed → verified (or needs-re-review)
 
     // Content (cached)
     contentHash: text('content_hash'),