Weekly AI Search Health Check (10 Minutes)

This week's material:
<homepage_html>
[PASTE HOMEPAGE HTML SOURCE]
</homepage_html>
<money_page_1_html>
[PASTE MONEY PAGE 1 HTML SOURCE]
</money_page_1_html>
<money_page_2_html>
[PASTE MONEY PAGE 2 HTML SOURCE]
</money_page_2_html>
<money_page_3_html>
[PASTE MONEY PAGE 3 HTML SOURCE]
</money_page_3_html>
<robots_txt>
[PASTE CURRENT ROBOTS.TXT]
</robots_txt>
<search_console>
[PASTE NEW "ENHANCEMENTS" ERRORS AND WARNINGS FROM THE LAST 7 DAYS, or "no new errors"]
</search_console>
<site_updates>
[CMS / THEME / PLUGIN UPDATES IN THE LAST 7 DAYS, ONE LINE EACH, or "none"]
</site_updates>

Act as my AEO health monitor. Run a 10-minute weekly check for the four regressions that quietly kill AI-search citations.

My site: homepage [https://YOUR-DOMAIN.COM] · top 3 money pages [URL1, URL2, URL3] · last week's scorecard from this prompt [PASTE, or BLANK on the first run]

Pasted above for the 4 URLs: 1. HTML source of each · 2. current robots.txt · 3. new Google Search Console "Enhancements" errors/warnings in the last 7 days (verbatim or "no new errors") · 4. CMS / theme / plugin updates in the last 7 days (one line each or "none").

Four checks, each marked ✅ PASS / ⚠️ ATTENTION / 🚨 REGRESSION:
Check 1 — Schema didn't disappear: per URL, count JSON-LD blocks and list each @type; compare to last week; a missing block = 🚨
Check 2 — robots.txt still allows AI crawlers: the AI search crawlers (OAI-SearchBot, PerplexityBot, Claude-SearchBot), the user-request agents (ChatGPT-User, Perplexity-User, Claude-User) and Bingbot are not blocked, where an agent the file does not name follows the * group and counts as not blocked unless that group blocks it; no new Disallow targeting them; sitemap directive present. Report the AI training controls (GPTBot, ClaudeBot, Google-Extended, Applebot-Extended) and the open archive crawler CCBot on their own line as blocked or not blocked, outside the status: blocking those is my choice, not a fault
Check 3 — Visible FAQs match FAQPage JSON-LD: for any URL with FAQPage, compare visible Q + first sentence of A to the JSON-LD; flag drift; flag any new microdata duplicating the JSON-LD
Check 4 — No new GSC Rich Results errors: per error/warning, the page, the schema field, and a one-line fix

Output:
1. This week's scorecard table: Check · URL · Status · One-line evidence
2. Diff vs last week — what changed, better or worse ("baseline established" on the first run)
3. This week's fix list, by urgency: what · why · time · which Unbought prompt from the list below — 🚨 this week, ⚠️ this month, ✅ leave alone
4. Save-this-scorecard block, copy-friendly for next week:
WEEK OF: [date]
HOMEPAGE — JSON-LD blocks: [list] — robots.txt allow-list: [PASS/ISSUES] — FAQ sync: [PASS/N/A/DRIFT] — GSC new errors: [N]
MONEY PAGE 1, 2, 3 — same fields

Unbought prompts you can point me to (Unbought is the prompt library this prompt comes from; use these exact titles, and if none fits, write "no matching prompt"):
- Write a robots.txt That Welcomes AI Crawlers
- Write Your Business's Organization Schema
- Convert Your Existing FAQ to FAQPage JSON-LD
- Turn Any How-To Blog Post into HowTo Schema
- Customer Review System That Generates Valid Review Schema
- Find Duplicate Schema Errors Killing Rich Results
- Explain Your Structured Data in Plain English

Success means every status cites the pasted source it came from, and a status is never ✅ for a URL whose inputs I didn't paste (mark it ⚠️ "not provided").

Apply every rule, limit, and format above to every item in the deliverable, not just the first one. Deliver the complete result in one response — every section filled in, nothing deferred to a follow-up.

If a required input is missing, ask for it once before starting; otherwise proceed and label any assumption you made.

Notes: R40 refresh; tuned to vendor prompting guidance