aeo-page-auditor

name: aeo-page-auditor description: > Audits any webpage or content against the AiVIS Cite Ledger ten-family (avs-v3) AI visibility scoring framework: Schema & Structured Data (18%), Entity & Heading Signals (14%), Authority & E-E-A-T (12%), Meta Tags & Open Graph (10%), Content Depth (10%), Crawlability & Bot Access (8%), Renderability & Page Speed (8%), Citation Signal Quality (8%), Indexability & Link Graph (6%), Security & Trust Signals (6%). Returns a 0–100 composite score, per-family grades (A–F), BRAG-style evidence-linked findings, and a prioritized fix list. Use this skill whenever the user pastes a URL, HTML, or page content and asks about AI visibility, citation readiness, AEO scoring, how AI engines see their page, why their content isn't being cited, or wants an audit. Also trigger when users ask "will ChatGPT cite this", "is my page AI-readable", or "what's my AiVIS Cite Ledger score".

AEO Page Auditor

Scores content against the AiVIS Cite Ledger ten-family (avs-v3) framework and returns evidence-linked findings with prioritized fixes.

Scoring formula

Composite = (Schema & Structured Data × 0.18) + (Entity & Heading Signals × 0.14) +
            (Authority & E-E-A-T × 0.12) + (Meta Tags & Open Graph × 0.10) +
            (Content Depth × 0.10) + (Crawlability & Bot Access × 0.08) +
            (Renderability & Page Speed × 0.08) + (Citation Signal Quality × 0.08) +
            (Indexability & Link Graph × 0.06) + (Security & Trust Signals × 0.06)

Each dimension is scored 0–100 independently before weighting. Hard-blocker rule: if Schema & Structured Data = 0, composite is capped at 52 regardless of other dimension scores.

Dimension scoring rubrics

Content Depth (18%)

Schema & Structured Data (20%)

AI Readability (12%)

Meta Tags & Open Graph (15%)

Heading Structure (10%)

Technical Trust (15%)

Score tiers

Output format

Always output in this exact structure:

## AiVIS Cite Ledger Audit, [Page title or URL]

**Composite score: [X]/100, [Grade] ([Tier name])**

### Dimension scores
| Dimension | Weight | Raw score | Weighted |
|---|---|---|---|
| Schema & Structured Data | 20% | [X]/100 | [X] |
| Content Depth | 18% | [X]/100 | [X] |
| Technical Trust | 15% | [X]/100 | [X] |
| Meta Tags & Open Graph | 15% | [X]/100 | [X] |
| AI Readability | 12% | [X]/100 | [X] |
| Heading Structure | 10% | [X]/100 | [X] |
| Security & Trust | 10% | [X]/100 | [X] |

### BRAG findings (high-confidence)
Each finding must include:
- **What**: specific issue observed
- **Evidence**: exact field or text from the page
- **Fix**: concrete implementation step
- **Score impact**: estimated point gain

### Priority fix order
1. [Highest weighted dimension gap first]
2. ...

### Citation readiness summary
One paragraph: what answer engines (ChatGPT, Perplexity, Gemini) would likely
do with this page in its current state.

BRAG evidence protocol

Every finding must pass four gates before inclusion:

1. Build from observed fields, finding must trace to a specific crawl-observable element
2. Reference explicit evidence, include the actual text/field, not a general statement
3. Audit recommendation linkage, fix maps directly to the finding it addresses
4. Ground in stored output, finding is verifiable by the user independently

Never include advisory findings (pattern-based guesses) without marking them [advisory]. High-confidence findings (directly observable) are unmarked and listed first.

Optimization loop guidance

Tell the user: a single audit is a diagnostic, not a solution. Recommend:

1. Fix the highest-weight dimension gap cluster
2. Re-audit
3. Compare category delta (not just overall score)
4. Repeat

Common failure mode to warn against: fixing all Security & Trust (10%) first while ignoring Schema & Structured Data (20%) and Content Depth (18%).

Reference files

• references/schema-patterns.md, Common JSON-LD patterns for each schema type
• references/dimension-examples.md, Before/after examples per dimension