Arguments: <url or local path> [--focus <comma-separated: crawlers,eeat,schema,passages,injection | all>] [--strict-mode]. Wherever <arguments> appears below, substitute the text the user typed after the skill name.
LLM SEO / Answer Engine Optimization Audit
Invokes the llm-seo-optimize agent to audit a site for answer-engine discoverability and citation-worthiness. Different from /seo-audit (traditional SERP ranking) -- this optimizes for getting quoted inside an LLM-generated answer.
CRITICAL RULES
- Delegate to the
llm-seo-optimize agent. This command is a thin wrapper -- the agent owns the full 6-phase audit.
- Verify live, not just code. If Playwright MCP is available, use it to confirm crawler access (
robots.txt live fetch), JSON-LD presence, and rendered passage extractability.
- Write output to
.aeo-audit/ for persistence so re-runs can diff against the baseline.
- Never fabricate crawler policies. If
robots.txt cannot be fetched, state the gap explicitly.
- Complementary to
/seo-audit, not a replacement. Traditional SEO and AEO require different optimizations -- flag the overlap, not duplication.
Procedure
1. Parse arguments
<url or local path>: required target -- live URL or path to static site / built HTML output--focus: restrict the audit to a subset of dimensions. Values are comma-combinable (--focus schema,eeat). Default: all, which runs every phase
crawlers -- the AI-bot robots.txt allowlist checkeeat -- E-E-A-T signal audit (authorship, citations, dates)schema -- JSON-LD / Schema.org structured data, plus citation readinesspassages -- passage-level extractability (direct-answer paragraphs, tables, bullet lists)injection -- prompt-injection hardening (hidden text, invisible CSS, comment payloads)
--strict-mode: report-level severity escalation. Every Warning is raised to Critical and the report opens with an explicit VERDICT: PASS / VERDICT: FAIL line, FAIL when any Critical remains
2. Spawn the agent
Invoke digital-marketing:llm-seo-optimize with the target, focus, and strict-mode flag. The agent loads its own knowledge base and runs the 6-phase protocol.
3. Report
Agent writes .aeo-audit/REPORT.md with:
# AEO Audit -- <site/page> -- <date>
## Summary
- Pages audited: N
- Blocked from: <list of engines with robots.txt denies>
- E-E-A-T score: X/5
- Extractability score: X/5
- JSON-LD coverage: P%
## Critical findings
- [CRITICAL] <page:line / selector> <issue>
## Per-page findings
...
## Cross-cutting recommendations
...
## Tracking setup
- Analytics filters for AI-referral hostnames: chatgpt.com (legacy chat.openai.com), perplexity.ai, claude.ai, copilot.microsoft.com, gemini.google.com
- Weekly brand-mention query set for citation-share tracking
Typical flow
# Full audit against a live URL
/llm-seo-audit https://example.com
# Focus on Schema + E-E-A-T only
/llm-seo-audit https://example.com --focus schema,eeat
# Static site audit
/llm-seo-audit ./dist/
# Strict grading: warnings become critical, report opens with PASS/FAIL
/llm-seo-audit https://example.com --strict-mode
Complementary commands
Run alongside traditional SEO tooling for a complete picture:
| Command |
Purpose |
Overlap with AEO |
/seo-audit |
Technical SEO (Core Web Vitals, meta tags, sitemap, redirects, SERP ranking) |
Low -- traditional ranking signals |
/content-strategy |
Conversion / CTA / tone / funnel audit |
Medium -- answer-worthiness often correlates with clarity |
/ga4-audit |
GA4 + GTM + Consent Mode v2 verification |
Measurement only -- AEO needs AI-referrer tracking set up in GA4 |
/llm-seo-audit |
AEO / answer-engine optimization |
-- |
Synergies
- Deep knowledge base ->
digital-marketing:llm-seo-optimize agent (the actual worker)
- Structured data validation alongside broader checks ->
/digital-marketing:seo-audit
- Measurement setup ->
digital-marketing:ga4-implementation-expert agent (AI-referrer tracking)
- Humanizing AI-sounding copy to raise E-E-A-T ->
/text-humanizer:humanize-text (text-humanizer plugin)
- Playwright-based live verification ->
playwright-skill
1---2name: llm-seo-audit3description: Checks whether a page is quotable by generative assistants, not merely rankable. TRIGGER WHEN: the user asks to audit for AI search or answer engines, check visibility in Google AI Overviews / SGE, Perplexity, ChatGPT Search, Claude Search, or Bing Copilot, verify the AI-bot crawler allowlist, or check E-E-A-T, llms.txt, JSON-LD citation readiness, passage extractability, and prompt-injection hardening. DO NOT TRIGGER WHEN: the task is traditional organic / SERP ranking SEO (use /digital-marketing:seo-audit), or content tone and voice only (use the content-marketer agent).4---56> Arguments: `<url or local path> [--focus <comma-separated: crawlers,eeat,schema,passages,injection | all>] [--strict-mode]`. Wherever `<arguments>` appears below, substitute the text the user typed after the skill name.78<!-- Generated by the Daodan compiler for codex. Edit the kernel, never this file. -->910# LLM SEO / Answer Engine Optimization Audit1112Invokes the `llm-seo-optimize` agent to audit a site for answer-engine discoverability and citation-worthiness. Different from `/seo-audit` (traditional SERP ranking) -- this optimizes for getting **quoted inside an LLM-generated answer**.1314## CRITICAL RULES15161. **Delegate to the `llm-seo-optimize` agent**. This command is a thin wrapper -- the agent owns the full 6-phase audit.172. **Verify live, not just code**. If Playwright MCP is available, use it to confirm crawler access (`robots.txt` live fetch), JSON-LD presence, and rendered passage extractability.183. **Write output to `.aeo-audit/` for persistence** so re-runs can diff against the baseline.194. **Never fabricate crawler policies**. If `robots.txt` cannot be fetched, state the gap explicitly.205. **Complementary to `/seo-audit`, not a replacement**. Traditional SEO and AEO require different optimizations -- flag the overlap, not duplication.2122## Procedure2324### 1. Parse arguments2526- `<url or local path>`: required target -- live URL or path to static site / built HTML output27- `--focus`: restrict the audit to a subset of dimensions. Values are comma-combinable (`--focus schema,eeat`). Default: `all`, which runs every phase28 - `crawlers` -- the AI-bot robots.txt allowlist check29 - `eeat` -- E-E-A-T signal audit (authorship, citations, dates)30 - `schema` -- JSON-LD / Schema.org structured data, plus citation readiness31 - `passages` -- passage-level extractability (direct-answer paragraphs, tables, bullet lists)32 - `injection` -- prompt-injection hardening (hidden text, invisible CSS, comment payloads)33- `--strict-mode`: report-level severity escalation. Every Warning is raised to Critical and the report opens with an explicit `VERDICT: PASS` / `VERDICT: FAIL` line, FAIL when any Critical remains3435### 2. Spawn the agent3637Invoke `digital-marketing:llm-seo-optimize` with the target, focus, and strict-mode flag. The agent loads its own knowledge base and runs the 6-phase protocol.3839### 3. Report4041Agent writes `.aeo-audit/REPORT.md` with:4243```44# AEO Audit -- <site/page> -- <date>4546## Summary47- Pages audited: N48- Blocked from: <list of engines with robots.txt denies>49- E-E-A-T score: X/550- Extractability score: X/551- JSON-LD coverage: P%5253## Critical findings54- [CRITICAL] <page:line / selector> <issue>5556## Per-page findings57...5859## Cross-cutting recommendations60...6162## Tracking setup63- Analytics filters for AI-referral hostnames: chatgpt.com (legacy chat.openai.com), perplexity.ai, claude.ai, copilot.microsoft.com, gemini.google.com64- Weekly brand-mention query set for citation-share tracking65```6667## Typical flow6869```70# Full audit against a live URL71/llm-seo-audit https://example.com7273# Focus on Schema + E-E-A-T only74/llm-seo-audit https://example.com --focus schema,eeat7576# Static site audit77/llm-seo-audit ./dist/7879# Strict grading: warnings become critical, report opens with PASS/FAIL80/llm-seo-audit https://example.com --strict-mode81```8283## Complementary commands8485Run alongside traditional SEO tooling for a complete picture:8687| Command | Purpose | Overlap with AEO |88|---------|---------|------------------|89| `/seo-audit` | Technical SEO (Core Web Vitals, meta tags, sitemap, redirects, SERP ranking) | Low -- traditional ranking signals |90| `/content-strategy` | Conversion / CTA / tone / funnel audit | Medium -- answer-worthiness often correlates with clarity |91| `/ga4-audit` | GA4 + GTM + Consent Mode v2 verification | Measurement only -- AEO needs AI-referrer tracking set up in GA4 |92| **`/llm-seo-audit`** | **AEO / answer-engine optimization** | -- |9394## Synergies9596- Deep knowledge base -> `digital-marketing:llm-seo-optimize` agent (the actual worker)97- Structured data validation alongside broader checks -> `/digital-marketing:seo-audit`98- Measurement setup -> `digital-marketing:ga4-implementation-expert` agent (AI-referrer tracking)99- Humanizing AI-sounding copy to raise E-E-A-T -> `/text-humanizer:humanize-text` (text-humanizer plugin)100- Playwright-based live verification -> `playwright-skill`
acaprino