diff --git a/agents/geo-analyzer.md b/agents/geo-analyzer.md
index 307e918..c3e9e3b 100644
--- a/agents/geo-analyzer.md
+++ b/agents/geo-analyzer.md
@@ -459,6 +459,12 @@ Load: `~/.claude/agents/resources/content-shape-for-ai.md`
 
 Sample 5-10 key pages (homepage + top service/blog pages). For each:
 
+**Record the denominator.** This samples; the report says "audit". Count the
+URLs in `sitemap.xml` for the coverage ratio, and carry it into the GEO
+SCORING block. No sitemap → total UNKNOWN, say so. Content shape is the
+axis most damaged by silent sampling: it is judged per page, so a 6-page
+sample of a 300-page site says nothing about the other 294.
+
 ### Checks
 
 1. **Definition Lead** — does the first sentence (or H1) follow
@@ -586,6 +592,7 @@ Score each axis. Use concrete findings from STEP 2-9.
 
 ```
 GEO SCORING (<depth>)
+COVERAGE                  : <N> of <M> sitemap URLs (<P>%) | <N> pages, total UNKNOWN
 AI Crawlers Policy        : XX/20  <justification>
 llms.txt                  : XX/20  <justification>
 Schema.org for AI         : XX/20  <justification>
@@ -596,6 +603,12 @@ AI Visibility (live)      : XX/20 | N/A (LOCAL)
 GEO GLOBAL (weighted)     : XX.X/20 (<depth>)
 ```
 
+**COVERAGE is mandatory, never omitted, never rounded up.** It bounds the
+per-page axes — Content Shape above all, and the page-level share of
+Schema.org. Site-wide axes (AI Crawlers Policy, llms.txt) are unaffected:
+robots.txt and llms.txt are single files, fully read. Say which is which
+rather than letting one ratio discredit the whole report.
+
 Per user instruction: **GEO weight in combined SEO+GEO report = 20% for
 local, 25% for national/SaaS/content.**
 
diff --git a/agents/seo-analyzer.md b/agents/seo-analyzer.md
index b00fdbd..4536e6e 100644
--- a/agents/seo-analyzer.md
+++ b/agents/seo-analyzer.md
@@ -400,8 +400,22 @@ Fetch rendered HTML. Extract and analyze:
 
 ## STEP 5 — ON-PAGE AUDIT `[both]`
 
+**Record the denominator BEFORE sampling.** This step samples; the report
+says "audit". Count the URLs in `sitemap.xml` (fetch it in full — the
+`head -50` in STEP 4 is a preview, not a count). That count is the coverage
+denominator, and it feeds the mandatory COVERAGE line in STEP 9. No sitemap
+→ denominator unknown: say so, never let silence imply full coverage. On a
+500-page site a 12-page sample is 2.4% — the On-page score is an
+extrapolation from it, and the reader cannot know that unless you print it.
+
 ### Meta tags per page (sample 5-15 key pages)
 
+Sample by risk, not convenience: homepage + top templates (one per page
+type: service, city, blog, product, legal) + any page GSC flags as a
+position 4-10 quick win. Same template audited twice buys nothing; an
+un-sampled template is an un-audited template — name the templates you
+skipped.
+
 For each sampled page:
 ```
 PAGE: <path>
@@ -738,6 +752,8 @@ misroutes the client-handover gate and the user's effort.
 
 ```
 SEO SCORING (<depth>)
+COVERAGE       : <N> of <M> sitemap URLs (<P>%) — templates skipped: <list|none>
+                 | <N> pages, total UNKNOWN (no sitemap)
 Technical      : XX/20  <justification>
 On-page        : XX/20  <justification>
 SEO Local      : XX/20 | N/A
@@ -749,6 +765,12 @@ Legal          : XX/20  <justification>
 SEO GLOBAL (weighted): XX.X/20 (<depth>)
 ```
 
+**COVERAGE is mandatory, never omitted, never rounded up.** It is the
+honesty bound on every page-level axis: On-page and the on-page share of
+Technical are extrapolations from the sample. If coverage < 25%, repeat it
+in §0 as a major alert — a 17/20 drawn from 3% of a site is not a 17/20, and
+`/client-handover` gates on these numbers.
+
 Per user instruction: this score represents **80% of the combined
 final score for local B2C (20% for GEO), or 75% for SaaS/national
 (25% for GEO)**. The `/seo` dispatcher combines SEO and GEO scores.