Content Audit

Content Audit Skill

Post-publication maintenance for AI search visibility. This skill audits published content for GEO compliance, detects freshness decay, tests LLM readability at the section level, and manages re-test schedules.

Lifecycle position: seo-geo (plan) → content-pipeline (create) → content-audit (maintain)

Reference Files — Load When Relevant

When This Skill Triggers

• "audit this page for GEO compliance"
• "check if the content is still fresh"
• "test the LLM readability of this section"
• "when should I re-test citation rates?"
• "is this section citation-ready?"
• "run a content maintenance check"
• "check for stale statistics"
• "are the opening sentences declarative enough?"

Core Workflow

1. Full Page Audit

When asked to audit a page, run all checks from references/compliance-checklist.md:

• Schema (Article + FAQPage + Organization + Review JSON-LD)
• Structure (single H1, TL;DR, FAQ, TOC, figure, key takeaways)
• Voice (first-person ≥2, "since 2004" in bio, no self-review)
• Links (internal ≥4, external ≥2, no homepage links)
• CTAs (closing CTA + ebook callout)
• Report PASS/FAIL per check with counts

2. Section-Level Quality Gate

When asked to review content quality, run all 5 questions from references/geo-quality-gate.md on every H2 section:

1. Direct declarative opener?
2. All pronouns resolved?
3. Single topic per section?
4. Verifiable data point present?
5. Citation-ready opening sentence?

3. Freshness Assessment

When asked about content freshness, use references/content-freshness.md:

• Identify every statistic and its source date
• Flag statistics >12 months old
• Apply the 4-priority framework (P1-P4)
• Generate a freshness update plan

4. LLM Readability Test

When asked about AI readability, use references/llm-readability.md:

• Run the 7-component self-test on each H2 section
• Report which components fail per section
• Identify the binding constraint on retrieval

5. Re-Test Cadence

When asked about re-testing, use references/retest-cadence.md:

• Check the 5 mandatory re-test triggers
• Recommend re-test schedule based on page age and content type
• Track citation rate baselines

Output Format

When auditing, produce a structured report:

# Content Audit: [Page Title]
**URL:** [url]
**Date:** [date]

## Compliance Score: X/25

| Category | Checks | Pass | Fail |
|----------|--------|------|------|
| Schema | 5 | X | X |
| Structure | 6 | X | X |
| Voice | 4 | X | X |
| Links | 3 | X | X |
| CTAs | 2 | X | X |
| New checks | 5 | X | X |

## Section Quality Gate

| Section (H2) | Opener | Pronouns | Topic | Data | Citation-ready | Result |
|---|---|---|---|---|---|---|
| [heading] | ✅/❌ | ✅/❌ | ✅/❌ | ✅/❌ | ✅/❌ | ✅/❌ |

## Freshness Assessment
[Priority level + specific statistics flagged]

## LLM Readability
[Component failures per section]

## Re-Test Status
[Current cadence + next scheduled re-test]

## Recommended Actions
1. [Highest priority fix]
2. [Second priority]
3. [etc.]

Rules

• Always check the RENDERED page (via curl or browser), not just post_content
• WordPress KSES may strip microdata — verify JSON-LD in page source
• Clear nginx cache before auditing: rm -rf /var/cache/nginx/fastcgi/* && systemctl reload nginx
• Never mark a page as compliant if any mandatory check fails
• Freshness checks are most critical for pages with time-sensitive queries