seo-audit

Use when running a full SEO optimization pipeline — technical audit, content analysis, keyword research, on-page optimization, structured data, AI citation visibility, performance deep dives, or schema validation. Also use when a site needs a comprehensive SEO review with an approval-gated action plan before execution.

Model	Source
sonnet	pack: seo

Full Reference

SEO Audit

Full SEO optimization pipeline — 11 phases, approval gate at phase 6. Reads every relevant file, pulls analytics data, audits technical health, checks AI citation visibility, evaluates content and schema, and produces a prioritized action plan. No execution happens without user sign-off.

Item	Value
Phases	8 (including 2.5, 4.5, 5.5 sub-phases)
Approval gate	Phase 6 — hard stop, no execution without sign-off
Delegated skills	ai-visibility, schema-markup, web-performance, lighthouse-api, google-search-console-api
Output	Prioritized action plan + optional site-report HTML deliverable

Reference Index

Usage: Read the reference file matching your current task from the index above. Each file is self-contained with steps and inline gotchas.

I want to…	File
Scan the codebase — build the page inventory	reference/phase1-codebase-inventory.md
Pull GSC data, CWV baselines, verify tracking	reference/phase2-analytics-data.md
Run the technical SEO checklist — tags, schema, crawl	reference/phase3-technical-audit.md
Run Lighthouse, escalate CWV failures	reference/phase4-performance.md
Analyze content quality, keyword gaps, schema audit	reference/phase5-content-keywords.md
Compile the report and present the approval gate	reference/phase6-report-approval.md
Execute approved items and verify after	reference/phase7-8-execution-verification.md

Phase Map

Phase	Focus	Output
1	Codebase scan	Page inventory table
2	Analytics pull	GSC + CWV data
2.5	AI citation check (invoke ai-visibility)	LLM visibility findings
3	Technical audit	Issue checklist
4	Lighthouse	Score table by page
4.5	Performance deep dive (invoke web-performance)	Diagnosis + fix plan
5	Content + keywords	Gap analysis
5.5	Structured data (invoke schema-markup)	JSON-LD audit
6	Report	Approval gate — STOP
7	Execution	Atomic fixes, commits
8	Verification	Re-run failing checks

Common Mistakes

Mistake	Correct approach
Skipping the approval gate	Always stop at Phase 6, always
Optimizing for keywords with no GSC data	Pull real data first — opinions are not rankings
Hardcoding aggregate ratings	Pull from actual review data or omit
Hiding content on mobile	Everything indexable must be visible on mobile
Title tags > 60 chars	Google truncates — keep tight
Duplicate meta descriptions	Every page needs a unique description
Skipping schema validation	Always validate — invalid schema does nothing
Committing all changes at once	Atomic commits — one fix per commit
Treating lab CWV as truth	Always compare against CrUX field data
Ignoring AI citation gap	LLM visibility is a real traffic channel now
Inline schema auditing	Delegate to schema-markup — it has the full spec

Related Skills

• google-search-console-api — GSC data pull
• lighthouse-api — performance audits
• crux-api — field data comparison
• ai-visibility — LLM citation check (Phase 2.5)
• web-performance — CWV diagnosis + fix plan (Phase 4.5)
• schema-markup — JSON-LD audit (Phase 5.5)
• server-side-tracking — analytics foundation verification
• local-seo-audit — location-specific SEO
• content-strategy — editorial planning from content gaps
• review-management — review volume and response strategy
• cro-audit — high-traffic low-conversion pages
• site-report — HTML deliverable generation

---

Announcement

┏━ 🔍 seo-audit ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓ ┃ Full SEO optimization pipeline — 11 phases, ┃ ┃ approval-gated action plan before execution ┃ ┗━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┛