Importance: High. Status: PENDING Sam's approval on execution lane. No live site or repo changes were made this session (diagnosis + verification only).
Triaged the full Sunday Site QA — 2026-06-07 log (547 scanned, 358 with findings, 525 findings). Pulled the complete findings list from Notion, then verified a representative sample of every issue type against the live DOM with Playwright (not just the top 15). Root-caused each category.
title_h1_mismatch (299 = 57% of all findings) is largely a miscalibrated scanner rule, not 299 broken pages. It compares the Rank Math SEO title vs the on-page H1 and flags low word-overlap. Verified:
These keyword-title vs conversion-H1 pairs are intentional and on-brand (sa-split-results-hero + copywriting skills). Mass-aligning H1s to titles would destroy the conversion heroes and brand voice. The category is chronic (236 → 237 → 299 over recent weeks) precisely because it is not a real defect. Recommendation: recalibrate the rule, do not bulk-edit.
.sa-hero wrapper (e.g. /locations/ashland/, /bardstown-... = 0; /locations/louisville/ = 18). Large batch.The scan set includes redirecting URLs and mislabels them: /assault-and-battery-...-louisville/ 301s to the homepage, inherits the home H1, and is logged as a title_h1_mismatch. The input list should be filtered to live 200s before comparison.
Diagnosis + verification complete. Held execution: the two highest-leverage actions need a deploy-lane decision. The scanner lives in the cloud sail-projects-root repo (GH Actions runner paths in the logs), not the local mirror. The live WP write needs a secret-safe lane (no WP MCP connected; the no-curl rule + Playwright code-echo make ad-hoc auth leak-prone) — best via the established wp-content-deployer / GH Actions path.