| name | indexing-audit |
| description | Audit indexing status with a site-wide overview first, then per-page deep-dive on top pages. Use when asked about crawling, indexing issues, or whether pages are indexed by Google. Aussi déclenché en français par "pourquoi mes pages sont pas indexées", "problème d'indexation", "mes pages sont pas dans Google", "Google ne crawle pas mon site", "combien de pages indexées", "découverte mais non indexée", "explorée mais non indexée", "est-ce que Google voit mes pages", "audit d'indexation". |
Indexing Audit
Audit the indexing health of a Google Search Console property: start with a site-wide count, then drill into the most-visible pages.
Steps
- Call
list_properties to confirm the exact site_url.
- Call
check_indexing_issues to get a site-wide breakdown of valid pages, errors, and warnings. Note any coverage categories with high error counts (this surfaces systemic issues before the per-URL drill-down).
- Call
get_search_analytics with dimensions=page, sort_by=impressions, row_limit=20 to identify the 20 most-visible pages.
- Extract the page URLs from step 3.
- Call
batch_url_inspection with up to 10 URLs at a time (API limit). Run twice if needed to cover all 20 pages.
- Categorize each URL by verdict:
- ✅ Indexed (PASS)
- ⚠️ Soft 404 / Excluded
- ❌ Not indexed / Blocked
- 🔍 Canonical mismatch (Google chose a different canonical)
- For each issue, record the specific
coverageState, pageFetchState, or robotsTxtState from the inspection.
Output format
Start with a site-wide summary from step 2: total valid / total errors / total warnings.
Then present a prioritized action list for the top pages:
- Critical: not indexed pages that have impressions (visibility being lost)
- High: canonical mismatches on high-traffic pages
- Medium: robots.txt or fetch blocks
- Low: soft exclusions on low-traffic pages
Include a summary table: Page URL | Verdict | Issue | Recommended action.
