AI Optimization Critic

JTBD (Jobs To Be Done)

Primary Job

When I have existing content and want to know whether it is eligible for Google AI Overviews, I want an evidence-backed AIO readiness review, so I can address the specific gaps before investing in Tier 2 content or Tier 3 platform work.

Secondary Jobs

When an ai-optimization-planner output needs independent validation before implementation, I want a critic review, so the plan is sound before resources are committed.
When a broader copy review needs an AIO eligibility lens, I want a focused perspective module, so the final recommendation includes AIO concerns without overclaiming.

This Skill Is For

Reviewing existing content for AI Overview eligibility gaps before publishing or refreshing.
Reviewing an ai-optimization-planner output before implementation begins.
Adding an AIO perspective to a copy-critic review on high-investment content.
Auditing pages after Tier 1 or Tier 2 implementation to verify AIO signal improvements.

This Skill Is NOT For

Writing or rewriting content — this skill is read-only.
Structural SEO review (heading hierarchy, meta descriptions, internal linking, intent alignment) — use seo-advisor.
Suggesting llms.txt, AI-specific markup, or keyword variants for AI modes — these are not AIO signals.
Predicting traffic or rankings.

Resolution Paths

User Situation	What Happens	What They Leave With
Content may be AIO-ineligible	Full 10-phase investigation	Verdict with CRITICAL/MAJOR/MINOR findings and actionable fixes
Plan needs validation before implementation	Critic reviews ai-optimization-planner output	ACCEPT or REVISE verdict with specific concerns
Broader review needs AIO lens	Perspective Module mode (abbreviated)	Compact AIO perspective for parent critic
Post-implementation check	Full review focused on implemented changes	Updated verdict confirming improvement or remaining gaps

When to Escalate

If content quality (tone, clarity, brand voice) is the primary problem → copy-critic
If structural SEO (headings, meta, internal links) is the primary problem → seo-advisor
If content doesn't exist yet → copy-planner + ai-optimization-planner
If REVISE verdict requires planning → ai-optimization-planner

Purpose

This critic is the companion reviewer to ai-optimization-planner. Its primary function is independent verification of AIO eligibility signals — ensuring that:

AIO-specific findings are genuinely about AIO, not structural SEO issues rebranded
E-E-A-T gaps are rated correctly against YMYL vs non-YMYL thresholds
Schema markup findings cite specific properties, not vague assertions
Fan-out coverage gaps identify the most commercially important absent angles
Scope boundary between AIO issues and seo-advisor/copy-critic issues is clearly maintained

The **Scope Boundary** output section is unique to this critic — it makes the domain boundary machine-parseable and prevents teams from conflating structural SEO with AIO eligibility.

What This Skill Does

✓ Predicts likely issues before investigation (pre-commitment, reducing confirmation bias)
✓ Verifies indexing status and snippet eligibility (RAG prerequisites)
✓ Assesses E-E-A-T across all four dimensions with documented severity calibration
✓ Generates and scores query fan-out coverage (5–7 related queries)
✓ Audits Schema.org markup with specific property-level evidence
✓ Reviews multimedia richness and commodity-content signals
✓ Applies Multi-Perspective Review (RAG system, AI consumer, agentic task, seo cross-check)
✓ Runs mandatory Self-Audit (confidence/refutability/flaw-vs-preference)
✓ Runs mandatory Realist Check (calibrate severity to documented impact)
✓ Reports structural SEO issues in a non-scoring **Scope Boundary** section
✓ Works as standalone full reviewer OR abbreviated perspective module

What This Skill Does NOT Do

✗ Write or edit content (read-only; disallowedTools: Write, Edit)
✗ Score structural SEO findings (heading hierarchy, meta descriptions, internal links)
✗ Recommend llms.txt or AI-specific markup
✗ Predict rankings or traffic
✗ Claim Core Web Vitals are direct AIO signals

Use_When

Content draft is ready and team wants AIO eligibility review before publishing
High-investment content (original research, case study, product page) needs AIO validation
Reviewing ai-optimization-planner output before Tier 1/2/3 implementation begins
Post-implementation verification (was schema markup added correctly? did E-E-A-T improve?)
Adding AIO perspective to a copy-critic review on strategic content

Do_Not_Use_When

Need structural SEO review (heading hierarchy, meta, internal links) → use seo-advisor
Need content quality review (tone, clarity, brand voice) → use copy-critic
Content doesn't exist yet → use copy-planner and ai-optimization-planner
You want to know if content "will rank" → not deterministic; no tool can answer this

Modes

Standalone Mode

When invoked with /ai-optimization-critic [content or URL]:

Run Phase 1: Pre-Commitment Predictions
Run Phase 2: RAG Retrievability Investigation
Run Phase 3: E-E-A-T Signal Audit
Run Phase 4: Query Fan-Out Coverage Assessment
Run Phase 5: Semantic HTML and Schema Markup Audit
Run Phase 6: Multimedia Richness and Non-Commodity Content
Run Phase 7: Multi-Perspective Review
Run Phase 8: Self-Audit (mandatory)
Run Phase 9: Realist Check (mandatory)
Run Phase 10: Synthesis
Output full format contract (see below)

Perspective Module Mode

When invoked by copy-critic, seo-advisor, or another skill with perspective: ai-optimization-critic:

Run abbreviated review (Phases 2–3, top 2 fan-out findings, top schema gap):

Output only:

**AIO Perspective:**
- RAG eligibility: [CONFIRMED/ASSUMED/BLOCKED] — [1 sentence]
- E-E-A-T: [STRONG/PARTIAL/WEAK] — [top gap if any, 1 sentence]
- Fan-out coverage: [X/Y angles] — [most important absent angle]
- Schema: [key type present or the most critical missing type]
- Key recommendation: [single most impactful AIO improvement]

Output Format Contract

These headings are load-bearing — eval harness and downstream parsers depend on exact names:

**VERDICT: [REJECT / REVISE / ACCEPT-WITH-RESERVATIONS / ACCEPT]**
**Overall Assessment**
**Pre-commitment Predictions**
**Critical Findings**
**Major Findings**
**Minor Findings**
**Scope Boundary**
**What's Missing**
**Multi-Perspective Notes**
**Verdict Justification**
**Open Questions (unscored)**

The **Scope Boundary** section is unique to this critic. It lists structural SEO issues or content quality issues found during review — acknowledged but explicitly not scored as AIO findings. This section maintains the domain boundary with seo-advisor and copy-critic.

Verdict Scale

REJECT: One or more CRITICAL findings (blocks AIO retrieval OR YMYL E-E-A-T absent), or YMYL content with multiple MAJOR findings
REVISE: One or more MAJOR findings without CRITICAL; improvements are specific and actionable
ACCEPT-WITH-RESERVATIONS: Only MINOR findings; no MAJOR or CRITICAL; notable items in What's Missing
ACCEPT: No MAJOR or CRITICAL findings; a clean AIO eligibility review

Severity Scale

CRITICAL: Demonstrably blocks AIO retrieval (nosnippet directive, content not indexed with strong evidence) OR YMYL content with total absence of any E-E-A-T dimension
MAJOR: Documented AIO signal gap that is addressable — missing schema on content with an obvious schema type, pure commodity content with ABSENT Experience signals, primary fan-out angle entirely absent, <main> landmark absent on a substantial site
MINOR: May improve AIO inclusion but not clearly deterministic — additional schema fields, PARTIAL fan-out coverage, original images recommended, dates in text but not <time datetime>
ENHANCEMENT: Best practice improvement with no documented AIO impact

Companion_Skills

ai-optimization-planner: Companion planner — invoke when REVISE or REJECT to redesign the improvement plan
seo-advisor: When **Scope Boundary** section indicates structural SEO needs dedicated review
copy-critic: When REVISE verdict or **Scope Boundary** indicates content quality needs improvement
copy-planner: When REVISE verdict includes MAJOR E-E-A-T content gaps requiring original writing

Version: 1.0.0 Last Updated: 2026-05-19 Perspective Module Compatible: Yes Standalone Ready: Yes

ai-optimization-critic

SKILL.md.claude/skills/ai-optimization-critic/SKILL.md