| name | audit-deep |
| version | 1.2 |
| description | Deep SEO & AEO maturity audit with GSC + Webflow data. Post-sale baseline that scores 4 dimensions across 5 maturity levels with data-backed evidence. Produces a client-ready report with quantified opportunities, engagement plan, and appendix data tables. Works with Webflow, WordPress, and other CMS platforms.
Triggers: audit deep, deep audit, post-sale audit, implementation audit.
Requires: GSC MCP server. Recommended: Webflow MCP server. Optional: PageSpeed MCP, Keywords Everywhere.
Workflow: Discover → Fetch → Check → Score → Report → Save.
Command: /audit:deep
|
Deep Audit Skill
Post-sale SEO and AEO maturity audit with full data access. Uses GSC search analytics, Webflow CMS data, and URL inspection to score 4 dimensions across 5 maturity levels. Produces a client-ready engagement baseline with quantified opportunities, a phased roadmap, and supporting data tables.
This report is designed as a professional deliverable — the foundation for an engagement plan. It frames findings as business opportunities with data-backed impact estimates, and ends with a clear engagement structure.
This skill is read-only — it never modifies Webflow content.
Prerequisites
- Required: Google Search Console MCP server
- Recommended: Webflow MCP server — enables CMS schema checks (TD1–TD3, TD5, CD3, CD5). If unavailable, the skill runs in GSC-only mode and skips Webflow-dependent checks. Works with any CMS (WordPress, Webflow, Squarespace, etc.).
- Optional: PageSpeed MCP (for Core Web Vitals — LCP, TBT, CLS, mobile score)
- Optional: Keywords Everywhere MCP server (for volume/intent enrichment)
Maturity Model Reference
See references/maturity-model.md for the full rubric. Same model as /audit but evaluated with deep signals and higher confidence ceiling (Medium–High).
Critical Guards
⚡ GUARD — GSC MCP unavailable:
- Stop: "Deep audit requires GSC. Connect GSC MCP and try again, or run
/audit {url} for a quick assessment."
⚡ GUARD — Webflow MCP unavailable:
- Warn: "Webflow MCP not connected. Running in GSC-only mode — CMS schema checks (field completeness, indexation cross-reference, template schema) will be skipped. Connect Webflow MCP for full coverage, or continue with reduced scope."
- Continue with GSC + web signals. Mark all Webflow-dependent checks as
N/A — GSC-only mode in the report.
⚡ GUARD — Low traffic (< 500 impressions in 90 days):
- Warn: "Low traffic site. Statistical confidence is reduced."
- Proceed with caveats.
⚡ GUARD — User requests abort:
- Confirm, exit cleanly, output any partial results.
Phase 0: DISCOVER
0.1 MCP Discovery
Search for tools BEFORE starting:
- GSC MCP (required): Search
+gsc search analytics. If missing → stop.
- Webflow MCP (recommended): Search
+webflow data cms. If missing → warn and continue in GSC-only mode.
- PageSpeed MCP (optional): Search
+pagespeed performance audit. If missing → note, will skip Core Web Vitals section.
- Keywords Everywhere (optional): Search
+keywords everywhere volume. If missing → note in report, proceed.
0.2 CMS Platform Detection
Before prompting for user input, detect the CMS platform from the homepage HTML. Fetch the homepage and look for:
| Signal | Platform |
|---|
/wp-content/uploads/ or wp-json in HTML | WordPress |
data-wf-site attribute or webflow.com script | Webflow |
cdn.shopify.com | Shopify |
static.wixstatic.com | Wix |
squarespace-cdn.com | Squarespace |
| No match | Unknown / Custom |
Record {platform} and use it to:
- Determine which CMS checks are applicable (see Phase 2 check tags)
- Adapt the report header data sources line
- If Webflow is not detected but Webflow MCP is connected → warn: "Webflow MCP is connected but this site does not appear to be Webflow. CMS checks will be skipped."
- If WordPress is detected but Webflow MCP is the only CMS MCP → automatically enter GSC-only mode without asking
0.3 User Input
Prompt for:
Client email (required): [email]
Then use MCP to discover:
- GSC properties accessible
- Webflow sites/workspaces accessible
- Analytics sources available
If multiple matches: present a numbered selection list and ask user to choose.
If unique: auto-select and confirm.
0.4 Set {domain}
Extract domain from the selected GSC property URL. Normalize: strip www., lowercase.
The {domain} is used for:
- Report save path:
./{domain}/reports/audit-deep-YYYY-MM-DD.md
- Latest pointer:
./{domain}/reports/latest-deep.md
- Activity log:
./{domain}/reports/activity-log.md
0.5 Review Activity Log
Check ./{domain}/reports/activity-log.md:
- If it exists: show recent activity summary (last 10 entries)
- Redundancy check: if
/audit:deep was run in the last 30 days → warn with date
- Note if a quick audit exists (
./{domain}/reports/latest-quick.md) — reference it for comparison
- If not found: proceed silently
0.6 Load Config
Load .claude/seo-copilot-config.json if it exists. Extract:
business.name → branded query detectionseo.competitors → competitor mentions in queriesaudience.primary → framing recommendations
If not found: proceed with defaults, note "Run /getting-started for personalized recommendations."
Phase 1: FETCH
1.1 GSC Data (90 days)
Search analytics:
- Page-level: all pages with impressions — URL, clicks, impressions, CTR, position
- Query-level: all queries with impressions > 5 — query, page, clicks, impressions, CTR, position
URL inspection (representative set):
- Top pages by impressions/clicks (top 20)
- Key template pages (one per CMS collection type)
- Index status, last crawl date, mobile usability
Coverage/indexing signals (if available via API surface).
Large dataset handling: If query+page data is very large (thousands of rows), do not read the full file into context. Instead, extract targeted analyses:
- Top 100 queries by impressions
- Top 50 queries by clicks
- Cannibalization candidates: queries where 2+ pages appear — use a targeted script (
jq or python3) rather than reading the raw file
- Sample first 200 rows for pattern analysis
Note any sampling in the report footnote.
1.2 Webflow Data [Webflow only — skip if GSC-only mode]
Pages and collections:
get_collection_list → list_collection_items per collection- For each CMS collection: fields, template settings, slug rules, SEO title/description mapping, Open Graph mapping
- All static pages: name, path, SEO title, meta description
CMS schema review per collection:
For each blog/article/case study collection, run the recommended schema check against the five groups defined in /cms-collection-setup:
| Group | Fields checked | Weight |
|---|
| Core | H1 override, Meta SEO title, Meta SEO description | 30% |
| Content | Main content (RichText), Main visual (Image), Alt text, Reading time, Updated date | 25% |
| Authority | Author (ref/multi-ref), Category, Tags, Pillar page | 20% |
| Enhancement | Primary keyword, Secondary keywords, Related posts, FAQ | 15% |
| Social | OG title, OG description, OG image | 10% |
Use alias matching — seo-title, meta-title, og-title-seo all count for Meta SEO Title, etc.
For each collection, compute:
- Per-group coverage score (0–100%)
- Overall weighted schema score
- List of missing fields by group
Surface schema scores in the Technical section of the report and in the CMS field completeness check (CD5). Collections scoring below 60% overall are a "Must fix" level finding.
Template HTML: Extract rendered HTML for key pages (one per collection type + homepage) to verify schema/tracking output matches config.
Sample first 100 items if a collection is large. Note limitation in report.
1.3 PageSpeed Data (Optional)
If PageSpeed MCP is available, run audits on:
- Homepage
- Top 2 pages by GSC impressions
Extract and record:
- Mobile performance score (0–100)
- Desktop performance score (0–100)
- LCP (Largest Contentful Paint)
- TBT (Total Blocking Time)
- CLS (Cumulative Layout Shift)
- Key diagnostics: render-blocking resources, image sizes, unused JS
Surface scores in the Technical section. Mobile score below 50 is a "Must fix" finding.
If PageSpeed MCP is unavailable, note: "Core Web Vitals data not available — connect PageSpeed MCP for performance diagnostics."
1.4 Crawl Tool Data (Optional)
Ask the user once:
"Do you have a Screaming Frog or Sitebulb export? If so, share the path to the issues overview CSV and I'll incorporate it."
Options: 1. Yes — provide path 2. No — skip
If provided, read the CSV and map these issue types to the existing check framework:
| Crawl Tool Issue | Maps to Check |
|---|
| H1: Missing | C2 / H1 coverage |
| Meta Description: Missing | TD7 / metadata completeness |
| Page Titles: Below 30 Characters | TD7 / metadata completeness |
| Images: Over 100 KB | Performance |
| Images: Missing Size Attributes | CLS / performance |
| Images: Missing Alt Text | Accessibility / SEO |
| Content: Lorem Ipsum Placeholder | Content quality |
| Directives: Noindex | TD1 / indexation |
| Links: Internal Nofollow Outlinks | Internal linking equity |
| Content: Low Content Pages | CD3 / thin content |
| Security: Missing X-Frame-Options/CSP/HSTS | Security (low SEO impact) |
When crawl data is available, it takes precedence over web-signal estimates. Note in the report: "Structural issues confirmed by [Tool] crawl of [N] pages."
1.5 Web Data
Fetch via WebFetch:
- Homepage HTML (for schema, tracking scripts, site-wide signals)
- sitemap.xml (try
/sitemap_index.xml as fallback)
- robots.txt
1.6 Keywords Everywhere (Optional)
If available: top 30 queries by impressions — volume, CPC, competition, trend.
Phase 2: CHECK
Run all quick audit checks (see /audit skill) on the fetched homepage HTML, PLUS the deep-mode checks below. Every check produces: Pass/Fail, Evidence, Source URL/template.
IMPORTANT: Check IDs (CD1, TD1, etc.) are for internal skill logic only. They must NEVER appear in the client-facing report. The report uses plain language descriptions and data evidence instead.
Content Checks (Deep)
| # | Check | Rule | Evidence | Requires |
|---|
| CD1 | Query-to-page alignment | For each top GSC query, a relevant page exists (query appears in page's title or description) | Unmatched queries = content gaps | GSC |
| CD2 | No cannibalization | No query cluster has 2+ pages ranking within 5 positions of each other | Conflicting pages + positions | GSC |
| CD3 | No thin content | All CMS items have body content ≥ 300 words (or ≥ 50% of fields populated) | Items below threshold | Webflow or crawl tool |
| CD4 | Topic coverage | GSC query clusters all have corresponding content pages | Uncovered clusters | GSC |
| CD5 | CMS field completeness | Per collection: % of items with all SEO fields populated (title, description, keywords) | Completion rate per collection | Webflow |
| CD6 | Editorial → product linking | For each editorial/blog page ranking in GSC, at least one contextual in-content link points to a relevant product/service page | Pages with no product link = missed conversion path | GSC + WebFetch |
Technical Checks (Deep)
| # | Check | Rule | Evidence | Requires |
|---|
| TD1 | Indexation cross-reference | All published pages appear in GSC | Not-indexed list, orphan-indexed list | Webflow |
| TD2 | Metadata completeness | All CMS templates have title/description rules, not falling back to name only | Templates using name as title | Webflow |
| TD3 | Schema coverage per template | Each content template type has appropriate schema (Article for blog, etc.) | Templates missing schema | Webflow |
| TD4 | Crawl freshness | Top pages crawled within last 30 days | Pages with stale crawl | GSC |
| TD5 | Sitemap vs live pages | Sitemap URL count matches published page count (within 10%) | Mismatch count | Webflow or sitemap fetch |
| TD6 | No duplicate meta titles | Zero duplicate <title> values across all pages | Duplicate groups | GSC or crawl tool |
| TD7 | No metadata length issues | All titles 30–60 chars, all descriptions 70–155 chars | Pages outside range | GSC or crawl tool |
| TD8 | Core Web Vitals | Mobile PageSpeed ≥ 50, LCP < 4s, CLS < 0.1 | Score, LCP, TBT, CLS values | PageSpeed MCP |
For checks marked "Webflow" in the Requires column: if Webflow MCP is unavailable, mark as N/A — GSC-only mode and note which signals were used as a partial proxy (e.g., sitemap URL count from WebFetch for TD5).
Authority Checks (Deep)
| # | Check | Rule | Evidence |
|---|
| AD1 | Branded query volume | Branded queries (containing business.name) exist and have clicks | Branded click/impression counts |
| AD2 | Branded trend positive | Branded query clicks not declining month-over-month | Trend direction |
| AD3 | Author attribution coverage | Author byline present on ≥ 80% of blog CMS items | Coverage % |
| AD4 | Author pages exist | CMS has author collection or about page with author details | Author entity evidence |
| AD5 | E-E-A-T consistency | Experience/credential markers consistent across CMS items | Consistency score |
Measurement Checks (Deep)
| # | Check | Rule | Evidence |
|---|
| MD1 | GSC data freshness | GSC has ≥ 90 days of data, most recent data within 3 days | Date range |
| MD2 | Tracking on all templates | Analytics script present on all rendered template pages | Templates missing tracking |
| MD3 | Existing reports | ./{domain}/reports/ contains weekly or monthly reports | Report files found |
| MD4 | Conversion tracking | GA4 events or goal configuration detected | Event patterns found |
Phase 3: SCORE
Same maturity model as /audit, evaluated with deep signals. Higher confidence ceiling.
Scoring Engine
Same level-by-level gate evaluation:
score_dimension(dimension, evidence):
for level in [1, 2, 3, 4, 5]:
if all gates for this level pass:
continue
else:
return level - 1 (or 1 if level 1 fails)
return 5
Content Dimension Gates (Deep)
| Level | Gates (all must pass) |
|---|
| 1 | Quick C1 (title) AND C2 (H1) AND GSC shows ≥ 1 ranking query per indexed page |
| 2 | Level 1 AND CD1 (query-page alignment for top queries) AND quick C4 (FAQ) AND CD5 (≥ 50% CMS field completeness — skip if GSC-only mode) |
| 3 | Level 2 AND CD2 (no major cannibalization) AND CD4 (topic coverage ≥ 80%) AND CD3 (no thin content) AND CD6 (editorial pages have product links) |
| 4 | Level 3 AND pillar pages rank for 50+ queries AND child pages link back AND CD5 (≥ 90% field completeness) |
| 5 | Level 4 AND automated freshness signals AND expanding query footprint (more queries month-over-month) |
Technical Dimension Gates (Deep)
| Level | Gates (all must pass) |
|---|
| 1 | Quick T1 (HTTPS) AND T2 (viewport) AND title tags exist on ≥ 80% of pages |
| 2 | Level 1 AND TD1 (all key pages indexed — skip if GSC-only) AND TD6 (no duplicate titles) AND TD5 (sitemap matches) |
| 3 | Level 2 AND TD3 (schema on all content templates — skip if GSC-only) AND TD2 (no templates using name as title — skip if GSC-only) AND TD7 (metadata lengths ok) AND TD8 (mobile PageSpeed ≥ 50 — skip if no PageSpeed MCP) |
| 4 | Level 3 AND Organization + Person schema AND BreadcrumbList AND TD4 (crawl freshness < 30 days for top pages) |
| 5 | Level 4 AND zero technical debt AND programmatic schema AND automated monitoring |
Authority Dimension Gates (Deep)
| Level | Gates (all must pass) |
|---|
| 1 | Always passes (baseline) |
| 2 | AD1 (branded queries exist) AND AD3 (author attribution ≥ 50%) |
| 3 | Level 2 AND AD2 (branded trend not declining) AND AD4 (author pages exist) AND AD5 (E-E-A-T consistent) |
| 4 | Level 3 AND high branded query ratio (≥ 20% of total queries) AND AI citation evidence |
| 5 | Level 4 AND dominant branded presence AND programmatic authority signals |
Measurement Dimension Gates (Deep)
| Level | Gates (all must pass) |
|---|
| 1 | Always passes (baseline) |
| 2 | MD1 (GSC ≥ 90 days) AND analytics installed on homepage |
| 3 | Level 2 AND MD2 (tracking on all templates) AND MD3 (existing reports found) |
| 4 | Level 3 AND MD4 (conversion tracking) AND LLM citation data available |
| 5 | Level 4 AND automated monitoring AND alert-driven optimization |
Overall Score
overall_level = min(content_score, technical_score, authority_score, measurement_score)
confidence = "Medium-High"
Prioritization (Impact/Confidence/Effort)
Score every finding as a recommendation:
| Factor | Scale | Criteria |
|---|
| Impact | 1-5 | Traffic at risk, % of total, business value |
| Confidence | 1-5 | Data volume, multi-signal confirmation |
| Effort | 1-5 | 1 = metadata fix, 2 = content tweak, 3 = section rewrite, 4 = new content, 5 = architecture change |
Priority = (Impact × Confidence) / Effort
| Bucket | Criteria | Action |
|---|
| Must fix | Priority ≥ 8.0 | Address immediately |
| High impact | Priority ≥ 4.0 | First sprint |
| Nice to have | Priority ≥ 1.5 | Backlog |
Phase 4: REPORT & SAVE
4.1 Report Principles
The report is a professional client deliverable — the foundation for an engagement. Follow these rules:
- No check IDs — Never write CD1, TD3, AD2, etc. in the report. Use plain language: "content gaps", "indexation coverage", "branded search trend".
- Opportunities, not failures — Frame findings as untapped potential with quantified upside. "You have X impressions going to pages without optimized titles — that's Y potential clicks per month" not "X pages have bad titles".
- No implementation details — Describe what needs attention and why it matters with data. Never describe how to fix it in Webflow/GSC. The how is the engagement.
- Quantify everything — Use GSC data to attach numbers: impressions, clicks, positions, trends. "500 monthly impressions with 0.8% CTR" is more persuasive than "low CTR".
- Business language — Revenue potential, competitive positioning, market visibility. Not technical jargon.
- Lead to engagement plan — The report should make the client confident that you understand their situation and have a plan.
4.2 Report Structure
Output a single markdown file with these sections:
Header:
# SEO & AEO Maturity Audit — {domain}
**Prepared for:** {domain}
**Date:** YYYY-MM-DD
**Assessment type:** Deep ({platform} + GSC data)
**Confidence:** Medium–High
**Period analyzed:** Last 90 days
**Data sources:** GSC ([X] days of data, [Y] pages, [Z] queries)[, Webflow ([N] pages, [M] CMS items across [K] collections) — omit if GSC-only mode][, PageSpeed (mobile [score]/100, desktop [score]/100) — omit if unavailable][, [Crawl Tool] crawl ([N] pages) — omit if not provided], Keywords Everywhere ([status])
**CMS platform:** {platform}
Adapt the data sources line to only list sources that were actually used.
1. About This Assessment
Brief methodology section (4-5 sentences):
- What data sources were analyzed and their coverage
- How scoring works (4 dimensions, 5 maturity levels, evidence-based with data backing)
- Confidence level and what it means
- How this relates to any prior quick audit (if
./{domain}/reports/latest-quick.md exists: "This builds on the discovery assessment from [date], now with full data access.")
2. Executive Summary
5-7 sentences for a decision-maker:
- Current maturity level and what it means for the business
- Total search footprint (pages indexed, queries ranking, total impressions/clicks)
- The single biggest opportunity with estimated impact (use GSC data)
- What's already strong (give credit — builds trust and shows fairness)
- Key risk if nothing changes (framed as competitive or visibility consequence)
- What reaching the next level would unlock in business terms
3. SEO & AEO Maturity Scorecard
Overall level (number + name) with a one-line business interpretation.
Table with NO check IDs:
| Dimension | Score | Level | What's Working | Biggest Opportunity |
|---|
| Content | X/5 | Name | [data-backed strength] | [quantified opportunity] |
| Technical | X/5 | Name | [data-backed strength] | [quantified opportunity] |
| Authority | X/5 | Name | [data-backed strength] | [quantified opportunity] |
| Measurement | X/5 | Name | [data-backed strength] | [quantified opportunity] |
Below the table:
- Strongest dimension and why it's a competitive advantage
- Weakest dimension and estimated business impact (use real numbers)
4. Key Findings
For each dimension, write a data-backed narrative (not a pass/fail table):
Content:
- Lead with strengths (total content pages, queries ranking, strong topics)
- Content gaps: queries with high impressions but no matching page (quantify with impressions)
- Cannibalization: where pages compete against each other (list specific pages + queries)
- Thin content: CMS items below quality thresholds
- CMS field completeness: what's populated vs. what's missing across collections
- End with: what a structured content strategy would unlock
Technical:
- Lead with strengths (indexation rate, schema coverage, clean URLs)
- Indexation gaps: published pages not appearing in GSC
- Metadata issues: duplicates, length problems, templates falling back to item name
- Schema gaps: templates missing appropriate structured data
- CMS schema coverage: for each content collection, report the schema score (Core/Content/Authority/Enhancement/Social) and overall %. Flag collections below 60% as a "Must fix" — they block
/refresh-content from populating key fields. Point to /cms-collection-setup:review to fix.
- End with: what technical optimization would unlock (more indexed pages = more visibility)
Authority:
- Lead with strengths (branded queries, author presence, E-E-A-T signals)
- Branded search trend (growing/stable/declining — with numbers)
- Author attribution coverage across content
- E-E-A-T consistency across the site
- End with: what stronger authority signals would mean for AI citations and rankings
Measurement:
- Lead with strengths (GSC data quality, analytics coverage)
- Tracking gaps across templates
- Reporting cadence and whether data-driven decisions are possible
- End with: what a measurement framework would enable
5. Opportunities & Business Impact
Prioritized table of opportunities. Use ICE scoring internally and show the score — it makes prioritization logic transparent and defensible in client conversations:
| Priority | ICE | Opportunity | Estimated Impact | What You're Leaving on the Table |
|---|
| Must fix | 9.0 | [opportunity] | [quantified: impressions, clicks, pages affected] | [business consequence] |
| High impact | 6.0 | [opportunity] | [quantified] | [business consequence] |
| Nice to have | 2.5 | [opportunity] | [quantified] | [business consequence] |
ICE = (Impact × Confidence) / Effort. Show one decimal place. Clients appreciate the rigor and it prevents scope arguments.
For each opportunity:
- Quantify with GSC data where possible (impressions at risk, potential click gain from CTR improvement, etc.)
- Frame as business outcomes (visibility, competitive position, AI readiness)
- Do NOT include how to fix — that's the engagement
6. Quick Wins
3-5 items that can show measurable results within the first 1-2 weeks. These build client confidence.
For each quick win:
- What it is (outcome, not implementation)
- Estimated impact (use data: "X pages affected, Y impressions at stake")
- Why it's quick (low effort, high confidence)
Frame as: "These are the first things we'd address — you'll see changes in search performance within 2-3 weeks."
7. Strategic Roadmap
Phased engagement plan:
Phase 1: Foundation (Weeks 1-2)
- Quick wins from Section 6
- Must-fix items from Section 5
- Expected outcome: [specific, measurable]
Phase 2: Optimization (Weeks 3-6)
- High-impact items from Section 5
- Content gap priorities
- Expected outcome: [specific, measurable]
Phase 3: Growth (Weeks 7-12)
- Nice-to-have items
- Advanced schema, authority building
- Expected outcome: [specific, measurable]
Ongoing: Monitoring
- Weekly performance tracking
- Monthly progress reviews
- Quarterly re-assessment
Include a maturity progression target:
Current: Level X → Phase 1 target: Level Y → Phase 3 target: Level Z
8. Engagement Recommendation
Professional proposal bridge:
"Based on this assessment, here's what we recommend:"
Recommended engagement:
- Scope: [what's included — reference the phases above]
- Approach: [quick wins first, then strategic, with progress tracking]
- Tracking: Weekly reports to measure impact, monthly reviews, quarterly deep audit re-run
- Expected outcomes: [specific maturity level targets, visibility improvements]
End with a confident closing line and clear next step (e.g., "Let's schedule a call to walk through the roadmap and align on priorities.").
Appendix A: Metadata Audit
Table (no check IDs, clean data):
| Page / Template | Title | Description | Open Graph | Issues |
|---|
| [page] | [title or rule] | [description or rule] | [status] | [plain language issue] |
Appendix B: Content Gaps
| Query Cluster | Monthly Impressions | Current Coverage | Suggested Action |
|---|
| [cluster] | [number] | No matching page / partial match | [create/expand/consolidate] |
Appendix C: Cannibalization
| Query Cluster | Competing Pages | Positions | Suggested Action |
|---|
| [cluster] | [page1, page2] | [pos1, pos2] | [consolidate/differentiate/redirect] |
Appendix D: Indexation Cross-Reference
| URL | In Webflow | In GSC | Status | Notes |
|---|
| [url] | Yes/No | Indexed/Not indexed | [status] | [plain language note] |
4.3 Save to File
Save these files:
./{domain}/reports/audit-deep-YYYY-MM-DD.md (timestamped)./{domain}/reports/latest-deep.md (overwrite each run — other skills read this)
Save supporting datasets as JSON under ./{domain}/reports/data/:
content-gaps.jsoncannibalization.jsonindexation.jsonmetadata-audit.jsonbaseline.json — structured snapshot for /weekly-report delta comparisons:
{
"domain": "{domain}",
"date": "YYYY-MM-DD",
"platform": "{platform}",
"maturity": {
"overall": 0,
"content": 0,
"technical": 0,
"authority": 0,
"measurement": 0
},
"search_footprint": {
"total_impressions_90d": 0,
"total_clicks_90d": 0,
"avg_ctr": 0.0,
"avg_position": 0.0,
"pages_with_impressions": 0,
"queries_with_impressions": 0
},
"top_pages": [
{"url": "", "impressions": 0, "clicks": 0, "ctr": 0.0, "position": 0.0}
],
"pagespeed": {
"mobile_score": null,
"desktop_score": null,
"lcp_mobile": null,
"tbt_mobile": null,
"cls_mobile": null
}
}
/weekly-report loads baseline.json to compute week-over-week deltas against this engagement starting point.
Create directories if needed: mkdir -p ./{domain}/reports/data/
4.4 Activity Log
Append to ./{domain}/reports/activity-log.md:
| YYYY-MM-DD | /audit:deep | Deep audit ({platform}). Overall Level: X. Content: X, Technical: X, Authority: X, Measurement: X. Must-fix: N. [Mobile PageSpeed: XX/100.] |
Log even on early exit.
Integration with Other Skills
| Finding | Skill | When |
|---|
| Low CTR / bad meta tags | /click-recovery | Quick metadata fixes |
| Outdated content | /refresh-content {url} | Full content refresh |
| Striking distance keywords (positions 4-30) | /keywords-opportunity:striking | Prioritized page 1-3 rankings with volume data |
| Content gaps / new keyword topics | /keywords-opportunity:discover | KE-powered new keyword discovery |
| Full keyword opportunity map | /keywords-opportunity | Striking distance + new discovery |
| CMS schema gaps (missing fields, low schema score) | /cms-collection-setup:review | Add missing fields to existing collections |
| No blog collection yet | /cms-collection-setup:create | Build a new collection with the full recommended schema |
| Low AEO scores on key pages | /aeo-optimize {url} | Add FAQ, schema, direct answers, question H2s per page |
| Missing config | /getting-started | First-time setup |
| Quick pre-sale assessment | /audit {url} | Before connecting MCP |
| Ongoing monitoring | /weekly-report | Track progress weekly |
| Monthly progress | /monthly-report | Month-end review |
Engagement workflow:
/audit {url} for pre-sale discovery/audit:deep post-sale for engagement baseline- Quick wins with
/click-recovery and /refresh-content
/weekly-report to track and demonstrate progress- Re-run
/audit:deep quarterly to measure maturity progression
