| name | Cornerstone-Content-Audit |
| type | playbook |
| description | Use when running a cornerstone content audit, pillar content review, or content architecture audit. End-to-end: discover pages, identify cornerstones, health audit, cannibalization check, and enhancement plan.
|
| version | 0.1.0 |
Cornerstone Content Audit
Also known as: top-down cornerstone audit.
Overview
Audits a site's most important content โ the 3โ7 cornerstone pages that should rank highest, attract links, and be cited by AI search. Identifies those pages, evaluates health across traditional and AI-era dimensions, detects cannibalization and pruning opportunities, and produces a prioritized enhancement plan.
Hub & Spoke is the content architecture โ one hub owns a topic cluster; spokes answer sub-intents and link up. Cornerstone is the quality standard for those pages (depth, intent, linking, fan-out, citability) โ not a separate architecture. Use this playbook when cornerstones are already known or nominated from a catalog.
Not this playbook: Messy site, accidental hubs, or GSC/GA4 triangulation before you trust the architecture โ Hub & Spoke Discovery & Recovery.
Architecture
| Step | Skill | Location |
|---|
| Phase 1 | Site Content Catalog | ../../skills/site-content-catalog/SKILL.md |
| Phase 3b | Fan-Out Coverage Analysis | ../../skills/fan-out-coverage-analysis/SKILL.md |
| Phase 3c | Information Gain Evaluator | ../../skills/information-gain-evaluator/SKILL.md |
| Phase 4a | SEO Cannibalization Audit | ../../skills/seo-cannibalization-audit/SKILL.md |
| Phase 5h | robots.txt Audit | ../../skills/robots-txt-audit/SKILL.md |
Read each linked SKILL.md before calling it. Follow its workflow exactly.
Input policy (non-negotiable)
Collect from the user:
- Domain or sitemap URL โ required
- Scope โ
quick scan (Phases 1โ2), single-page cornerstone review (Phase 3 for one URL), or full audit (Phases 1โ6)
- Known cornerstones (optional) โ skip Phase 2 nomination when provided
- Optional context โ protected pages, topic clusters, architecture notes (never from workspace or examples)
If domain/sitemap is missing, ask once. Do not invent URLs, metrics, or recommendations before tool results exist.
First response rules
On the first user-facing response:
- Name the scope (
quick scan | single-page | full audit).
- Name exact skills for that scope (see Architecture).
- Preview deliverable shape for that scope.
- State the Phase 2 approval checkpoint when nomination is involved.
- Frame recommendations around audit dimensions (rankings, fan-out, information gain, AI citability, cannibalization, pruning, AI crawler permissions via
robots.txt).
Use AskQuestion for Phase 2 approval when available; otherwise wait for explicit user confirmation.
Phase 1: Discovery & catalog
Call: Site Content Catalog
Pass sitemap URL or domain. Receive handoff JSON v1.0 (REFERENCE.md in catalog skill). Read limitations[] โ v1.0 does not include inbound internal link counts.
If the catalog returns >200 pages, scope down with the user (path prefix, page type, or top traffic).
Output: Validated page catalog ready for cornerstone scoring.
Phase 2: Cornerstone identification
Goal: 3โ7 cornerstones grouped by topic cluster.
Step 2a: Topic clusters
From the Phase 1 catalog:
- Group by URL path, title themes, keyword clusters
- Target 3โ7 distinct topic areas with a clear head topic each
Step 2b: Score candidates
Per cluster, score top 3โ5 candidates:
| Signal | Source | Weight |
|---|
| Keyword authority | dataforseo_labs_google_ranked_keywords โ volume sum for top-20 ranks | High |
| Estimated traffic value | Same โ ETV | High |
| Backlink authority | backlinks_summary โ referring domains | High |
| Internal link gravity | Catalog โ inbound internal links | Medium |
| Content depth | Catalog โ length, headings | Medium |
| Navigation prominence | Top nav? Clicks from homepage? | Medium |
| Freshness | Catalog last modified | Low |
Step 2c: Nominate
Per cluster, nominate highest scorer. Flag flat scores, wrong-page cornerstones (blog beats LP), or competing pages.
Human checkpoint
Present nominations with URL, score, and rationale. Do not proceed to Phase 3 until the user approves or adjusts.
Output: Confirmed 3โ7 cornerstones with topic clusters and primary keywords.
Phase 3: Health audit
Run for each confirmed cornerstone (parallel OK).
3a: Traditional SEO scorecard
| Dimension | Source |
|---|
| Rankings | dataforseo_labs_google_ranked_keywords |
| Backlinks | backlinks_summary |
| Internal links | Phase 1 catalog |
| Content | on_page_content_parsing or WebFetch |
| SERP features | serp_organic_live_advanced |
3b: Fan-out coverage
Call: Fan-Out Coverage Analysis
Pass: keyword, domain, anchor_url (cornerstone URL).
3c: Information gain
Call: Information Gain Evaluator
Pass: target_url, primary_keyword.
3d: AI citability
ai_opt_llm_ment_search โ domain in LLM mentions for primary keyword
ai_opt_llm_ment_top_pages โ pages AI cites for keyword
- Structure check (WebFetch): H2/H3 hierarchy, data tables, FAQ blocks, extractable passages
- Crawlability flag (lightweight): if
robots.txt is already fetched in-session, note whether this cornerstone path is blocked for Googlebot or major AI answer bots โ full policy audit remains Phase 5h
3e: Scorecard template
## Scorecard: {cornerstone URL}
**Topic cluster:** {name}
**Primary keyword:** {kw} | Rank: #{pos} | Volume: {vol}/mo
### Traditional SEO
- Rankings: {top-20 count, ETV}
- Backlinks: {referring domains}
- Internal links: {inbound, outbound}
- Content: {words, headings, last updated}
- SERP features: {snippet, PAA, sitelinks}
### Fan-Out Coverage: {X}% ({covered}/{total})
### Information Gain: {High/Medium/Low}
### AI Citability: {mentions, competing pages, structure}
Phase 4: Cannibalization & pruning
4a: Cannibalization check
Call: SEO Cannibalization Audit
Pass cornerstone URLs, primary keywords, domain, and optional protected pages from user context.
4b: Pruning candidates
From Phase 1 catalog, flag pages with:
- Zero ranked keywords
- Near-zero backlinks
- Thin content (<300 words)
- Topical overlap with a cornerstone
Classify: Consolidate | Improve | Deindex | Delete
Output: Cannibalization conflicts + pruning list.
Phase 5: Enhancement plan
Synthesize Phases 3โ4. Per cornerstone: content (from 3c), internal links (from 3b), keyword realignment, SERP features, fan-out gap content, cannibalization fixes, pruning.
5h: robots.txt and AI crawler policy
Call: robots.txt Audit
Pass:
domain
key_pages โ confirmed cornerstone URLs from Phase 2
crawl_policy โ default max_discovery unless user specifies otherwise
Receive handoff JSON v1.0 (REFERENCE.md in robots skill) for the report's crawler policy section.
Cross-check when available:
- Phase 1 catalog โ sitemap URLs blocked by
Disallow
- Phase 3d โ weak AI mentions plus blocked answer bots โ flag as likely discoverability issue
- Phase 4 pruning โ pages marked Deindex vs training-crawler exposure (note only; do not block audit)
Prioritization
- Quick wins: internal links, title/meta/H1, schema, robots.txt fixes that unblock cornerstones or wrongly block discovery crawlers
- Medium effort: freshness, information gain, heading structure, clear pruning
- Strategic: new cluster content, major rewrites, architecture changes
Phase 6: Report
# Cornerstone Content Audit โ {domain} โ {date}
## Executive Summary
## Topic Cluster Map
## Cornerstone Scorecards
## Cannibalization & Pruning
## Enhancement Plan
### Quick Wins / Medium Effort / Strategic Projects
## New Content Opportunities
## robots.txt & AI Crawler Policy
## Appendix: Full Page Catalog
Scope options
| User says | Phases |
|---|
| Quick scan / identify cornerstones | 1โ2 |
| Audit {URL} | 3 only |
| Full audit | 1โ6 |
| Fan-out for {keyword} | Fan-out skill only |
| Information gain on {URL} | Information gain skill only |
When to use which playbook
DataForSEO tools used
dataforseo_labs_google_keywords_for_site, dataforseo_labs_google_ranked_keywords, dataforseo_labs_google_relevant_pages, dataforseo_labs_google_related_keywords, dataforseo_labs_google_keyword_suggestions, dataforseo_labs_google_keyword_ideas, backlinks_summary, serp_organic_live_advanced, on_page_content_parsing, ai_opt_llm_ment_search, ai_opt_llm_ment_top_pages
Research basis
Cornerstone content practice (Yoast, HubSpot, etc.) plus 2025โ2026 GEO: AI Overview fan-out, information gain for citations, selective robots.txt crawler policy (RFC 9309; training vs answer/search bots), content pruning impact, structured-content citation lift.
Requirements: REQUIREMENTS.md