claude-three-o

Claude Three-O - SEO + GEO + AAO Unified Optimization Plugin

Unified search and AI visibility optimization plugin for Claude Code. Three pillars — SEO, GEO, and AAO — analyzed in parallel with a single score (0-100) and prioritized action plan.

    

> Three-O = SEO (Search Engine) + GEO (Generative Engine) + AAO (Assistive Agent) > > Built for the Korean market by JaeHyunGround

한국어 README

Table of Contents

• Installation
• Quick Start
• Commands
• Scoring System
• Architecture
• Output Formats
• Analysis Accuracy
• Requirements

Installation

Plugin Install (Claude Code 1.0.33+)

/plugin install claude-three-o

Manual Install

git clone https://github.com/JaeHyunGround/claude-three-o.git
cd claude-three-o && bash install.sh

Quick Start

# Full 3-pillar audit
/three-o audit https://example.com

# Individual pillar
/three-o seo audit https://example.com
/three-o geo audit "Brand Name"
/three-o aao audit https://example.com

# Generate PDF report
/three-o report full --format pdf

Commands

Main Commands

| Command | Description | |---------|-------------| | /three-o audit <url> | Full 3-pillar audit (SEO + GEO + AAO parallel) | | /three-o report [type] | Unified report (md / json / pdf) | | /three-o plan <business> | Strategic optimization plan with timeline | | /three-o competitor <u1> <u2> | Cross-pillar competitor benchmarking | | /three-o dashboard | Dashboard data export | | /three-o setup | API key configuration wizard |

SEO Module

| Command | Description | |---------|-------------| | /three-o seo audit <url> | Full SEO analysis | | /three-o seo technical <url> | Technical SEO (meta, security, mobile, speed) | | /three-o seo naver <url> | Naver-specific SEO (Smart Store, Place) | | /three-o seo content <url> | Content quality and E-E-A-T signals | | /three-o seo schema <url> | Schema.org structured data audit | | /three-o seo schema generate <url> | Auto-generate JSON-LD from page content | | /three-o seo images <url> | Image optimization analysis | | /three-o seo drift baseline <url> | Capture SEO baseline for monitoring | | /three-o seo drift compare <url> | Detect changes from baseline |

GEO Module

| Command | Description | |---------|-------------| | /three-o geo audit <brand> | Full GEO analysis (all dimensions) | | /three-o geo mentions <brand> | AI platform mention tracking | | /three-o geo context <brand> | Context quality and sentiment analysis | | /three-o geo citability <url> | Passage-level AI citation readiness | | /three-o geo entity <brand> | Knowledge graph entity presence | | /three-o geo visibility <brand> | Position ranking in AI responses | | /three-o geo technical <url> | AI crawler accessibility | | /three-o geo llms-txt <url> | llms.txt validation and generation | | /three-o geo platforms <brand> | Platform-specific analysis (ChatGPT, Perplexity, Gemini, Claude) | | /three-o geo drift <brand> | GEO drift detection over time | | /three-o geo rewrite <url> | Content rewrite suggestions for AI citability |

AAO Module

| Command | Description | |---------|-------------| | /three-o aao audit <url> | Full AAO analysis | | /three-o aao selectability <url> | Agent selectability signals (6 dimensions) | | /three-o aao conversion <url> | Conversion funnel analysis | | /three-o aao data <url> | Structured data with action schema detection | | /three-o aao rendering <url> | SSR, JS dependency, semantic HTML | | /three-o aao entity <url> | NAP consistency + sameAs linking | | /three-o aao feed <url> | Product feed validation (Google Merchant / Naver EP) | | /three-o aao scenario <url> | Agent scenario testing by industry |

Scoring System

Three-O Score (0-100)

Three-O Score = (SEO × 35% + GEO × 35% + AAO × 30%) × balance_penalty

The balance penalty (0.85–1.0) penalizes imbalanced profiles — a site with SEO 90 but GEO 10 scores lower than pure weighted average. Confidence score (0–1.0) indicates data availability.

| Grade | Score | Meaning | |-------|-------|---------| | A+ | 90-100 | Excellent across all pillars | | A | 80-89 | Strong with minor gaps | | B+ | 70-79 | Good foundation, room for improvement | | B | 60-69 | Average, significant gaps | | C+ | 50-59 | Below average, action needed | | C | 40-49 | Poor, major issues | | D | 20-39 | Very poor visibility | | F | 0-19 | Minimal presence |

SEO Score

Technical (meta quality + security + mobile + headings + images + performance) + Content + On-Page + Schema + AI Readiness

Key features:

• Meta quality evaluation: not just presence but optimal length (title 30-60, desc 120-160), protocol consistency, duplication detection
• Heading hierarchy validation: single H1, proper H2→H3 nesting
• Image alt coverage: percentage-based scoring
• Weighted section scoring: meta_quality 30%, security 15%, mobile 15%, headings 15%, images 10%, performance 15%

GEO Score (Geometric Mean)

GEO = geometric_mean(MF^0.30 x CQ^0.25 x VR^0.20 x EP^0.15 x TA^0.10)

• MF (30%): Mention Frequency across AI platforms
• CQ (25%): Context Quality and accuracy
• VR (20%): Visibility Ranking position
• EP (15%): Entity Presence in knowledge graphs
• TA (10%): Technical Accessibility for AI crawlers

Partial scoring: Dimensions requiring API keys (MF, VR) are excluded when unavailable — remaining dimensions' weights are redistributed. Confidence score tracks data completeness.

Platform-specific GEO: Each AI platform (ChatGPT, Perplexity, Gemini, Claude) gets an individual GEO score based on its citation preferences:

• ChatGPT: definition sentences, concise paragraphs (80-300 chars), FAQ, freshness
• Perplexity: source attribution, data density, recency, outbound references
• Gemini: E-E-A-T signals, tables, comparison content, knowledge graph connectivity
• Claude: content depth, data-backed claims, nuance, research references

AAO Score

Selectability (industry-weighted) + Conversion Readiness + Structured Data + Rendering + Entity Consistency

Selectability dimensions: structured_data (25%) + reviews_ratings (20%) + info_completeness (20%) + api_booking (15%) + trust_signals (10%) + freshness (10%)

Key features:

• Industry auto-detection: restaurant, ecommerce, clinic, hotel, education, saas → weight adjustment
• Signal correlation: cross-dimension bonuses (e.g., schema + reviews = +8) and penalties
• Cross-validation: schema claims verified against actual page content
• Quality scoring: not just presence/absence, but quality assessment per field

Industry-Specific Adjustments (Korean Market)

| Industry | Adjustment | |----------|-----------| | E-commerce (Smart Store) | AAO +10% | | Franchise HQ | GEO +5%, AAO +5% | | Academy / Education | GEO +10% | | Clinic / Healthcare | GEO +10% | | Restaurant | SEO +5% (Naver Place) |

Architecture

claude-three-o/
  .claude-plugin/
    plugin.json              # Plugin manifest
    marketplace.json         # Marketplace catalog
  skills/                    # 32 skills
    three-o/                 # Main orchestrator
    seo-*/                   # SEO module (10 skills)
    geo-*/                   # GEO module (9 skills)
    aao-*/                   # AAO module (8 skills)
    three-o-*/               # Cross-cutting (4 skills)
  agents/                    # 24 subagents
  hooks/                     # Quality gate hooks (3)
  scripts/                   # 45 Python scripts
  schema/                    # Schema.org JSON-LD templates (10 industry types)
  tests/                     # Test suite (2431 tests)
  reports/                   # Generated reports (gitignored)

Quality Gate Hooks

| Hook | Purpose | |------|---------| | validate_quality.py | INP (not FID), no HowTo, FAQ gov/health only, no hardcoded paths | | check_cwv.py | Core Web Vitals terminology validation | | check_schema.py | Schema recommendation rules enforcement |

Output Formats

| Format | Command | Use Case | |--------|---------|----------| | Terminal | (default) | Quick analysis with color-coded bars | | JSON | --json | API integration, data pipeline | | Markdown | --format md | Documentation, PR descriptions | | PDF | --format pdf | Client deliverables, presentations |

PDF Report Features

• Korean text support (AppleGothic / NanumGothic)
• Color-coded score visualizations
• 6-page structure: Title, Executive Summary, SEO, GEO, AAO, Action Plan
• Priority-ordered action table (P0/P1/P2)
• JaeHyunGround branding

Analysis Accuracy

Three-O prioritizes realistic scoring over inflated numbers. Key accuracy mechanisms:

| Module | Mechanism | Effect | |--------|-----------|--------| | GEO Citability | 7-dimension citability scoring | Passage clarity, factual density, citation pattern, self-containment, quote readiness, structural format, authority signals | | GEO Citability | Sentence structure analysis | Subject-predicate detection, weak opener penalty, clear subject bonus | | GEO Citability | Weighted factual density | Numbers, proper nouns, technical terms, units scored by type and frequency | | GEO Citability | Quote readiness scoring | Quotable length, definitional pattern, standalone clarity, specificity | | GEO Citability | Platform-specific citation modeling | ChatGPT, Perplexity, Gemini, Claude preference weights and citation styles | | GEO Entity | 4-dimension entity scoring | Schema presence, connection strength, attribute completeness, disambiguation | | GEO Entity | Tiered sameAs verification | Knowledge (Wikidata/Wikipedia) > Authority (LinkedIn/Naver) > Social platforms | | GEO Entity | Type-aware attribute completeness | Per-entity-type required fields (Restaurant ≠ Organization ≠ Hotel) | | GEO Entity | Disambiguation signal detection | Business registration, founding year, CEO, location qualifier, generic name penalty | | GEO Sentiment | 5-dimension sentiment scoring | Polarity strength, consistency, coverage, platform alignment, signal diversity | | GEO Sentiment | Bilingual lexicon analysis | Korean (23 terms) + English (31 terms) sentiment lexicons with intensity weights | | GEO Sentiment | Cross-platform sentiment tracking | Per-platform score aggregation with alignment measurement | | GEO Sentiment | Confidence scoring | Data volume + platform diversity weighted confidence (0-1.0) | | GEO Platforms | 5-dimension per-platform scoring | Each platform scored across extractability, data density, E-E-A-T, depth, access | | GEO Platforms | Platform-specific citation modeling | ChatGPT (definitions, Q&A), Perplexity (facts, sources), Gemini (E-E-A-T, schema), Claude (nuance, evidence) | | GEO Platforms | Crawler access analysis | GPTBot, PerplexityBot, Google-Extended, ClaudeBot blocking detection | | GEO Platforms | Cross-platform gap analysis | Optimization imbalance detection with severity and priority actions | | GEO Platforms | Dynamic recency detection | Current year awareness, date metadata, schema dates, freshness language | | SEO Content | 4-axis E-E-A-T quality scoring | Experience, Expertise, Authoritativeness, Trust scored independently (0-100) | | SEO Content | Korean + English signal detection | Bilingual regex patterns per axis (5+6+7+8 signal categories) | | SEO Content | Weighted content score | depth_score × 0.4 + eeat_score × 0.6 with axis-level diagnostics | | SEO Technical | 8-dimension technical scoring | Meta quality, heading structure, image optimization, link health, mobile readiness, indexability, security, performance | | SEO Technical | Mobile readiness scoring | Viewport quality, responsive breakpoints, touch targets, click-to-call, flex/grid layout | | SEO Technical | Indexability analysis | Robots directives, canonical validation, hreflang, lang attribute, JSON-LD presence | | SEO Technical | Security signal detection | HTTPS, CSP meta, mixed content, SRI, inline event handler penalty, form action validation | | SEO Technical | Performance signal analysis | Resource hints, async/defer optimization, critical CSS, font-display, lazy loading | | SEO Technical | Image optimization scoring | Alt coverage, lazy/eager loading, srcset, modern formats (WebP/AVIF), explicit sizing | | SEO Technical | Link health assessment | Anchor diversity, generic anchor penalty, nofollow ratio, navigation quality, breadcrumb | | AAO Conversion | 6-dimension funnel scoring | CTA quality, form accessibility, flow completeness, mobile, deep link, confirmation | | AAO Conversion | CTA quality assessment | Specific vs generic CTA detection, schema action, urgency, ARIA roles | | AAO Conversion | Mobile conversion optimization | Viewport, click-to-call/map, touch targets, SNS login, sticky CTA | | AAO Conversion | Deep link & app integration | iOS universal links, Android intents, custom schemes, API endpoints | | AAO Conversion | Payment path depth analysis | Price visibility, payment methods, secure checkout, refund policy | | AAO Rendering | 6-dimension rendering scoring | SSR quality, JS dependency, semantic structure, content accessibility, agent crawlability, rendering resilience | | AAO Rendering | Hydration pattern detection | Static, SSR+hydrate, CSR, hybrid classification with framework awareness | | AAO Rendering | Framework SSR capability assessment | Next.js/Nuxt.js/Remix/Astro/Gatsby (SSR-capable) vs React/Vue/Angular (CSR-only) | | AAO Rendering | Agent crawlability scoring | Navigation schema, breadcrumb, canonical, robots, internal link structure | | AAO Rendering | Rendering resilience analysis | Noscript fallback, progressive enhancement, image loading strategy, resource hints | | AAO Feed | 6-dimension feed scoring | Data quality, field completeness, feed freshness, pricing accuracy, media quality, platform compliance | | AAO Feed | Google Merchant compliance | GTIN/MPN identifiers, availability/condition validation, category, shipping, custom labels | | AAO Feed | Naver Shopping compliance | Category depth, price_pc, shipping, manufacture/brand, origin, import_flag, event_words | | AAO Feed | Pricing accuracy analysis | Currency format, sale price consistency, zero price detection, effective dates, shipping | | AAO Feed | Media quality scoring | HTTPS validation, modern formats (WebP/AVIF), additional images, high-res indicators, alt text | | AAO Feed | Data quality assessment | Title length/uniqueness, description quality, link validation, ID uniqueness, HTML markup penalty | | AAO Selectability | Industry auto-detection + weight adjustment | Restaurant ≠ SaaS ≠ Clinic scoring | | AAO Selectability | Signal correlation bonuses/penalties | Schema + reviews = synergy bonus | | AAO Selectability | Cross-validation | Schema claims verified against page content | | Three-O Score | Balance penalty | Imbalanced pillar profiles penalized (0.85-1.0) | | GEO Score | Partial dimension support | Unavailable dimensions excluded, weights redistributed | | Drift Detection | SEO dimension-level tracking | Meta, headings, images, schema, performance changes with severity | | Drift Detection | Velocity + acceleration | Score change rate per snapshot with trend classification | | Drift Detection | Cross-pillar correlation alerts | Multi-pillar decline, single-pillar isolation, divergence detection | | Competitor Benchmark | Multi-dimensional gap analysis | Per-dimension scoring across 16 dimensions (SEO 5 + GEO 5 + AAO 6) | | Competitor Benchmark | Prioritized action plan | P0/P1/maintain based on gap severity with Korean recommendations | | Competitor Benchmark | Competitive positioning | Radar chart data with market averages per dimension | | Content Rewrite | Weakness analysis per passage | Detects vague language, filler, weak openers, low factual density | | Content Rewrite | Strategy-based rewrite suggestions | Definition, comparison, data, causal patterns with templates | | Content Rewrite | Platform-specific tips | Per-platform (ChatGPT, Perplexity, Gemini, Claude) optimization tips | | Schema Generator | Industry auto-detection + template selection | Restaurant, clinic, hotel, ecommerce, education, SaaS, agency, realestate templates | | Schema Generator | HTML content extraction | Extracts name, phone, address, hours, prices, ratings, social links, services, email, logo, parent org | | Schema Generator | Coverage scoring + suggestions | Identifies missing fields with impact-based improvement tips | | All Scores | Confidence tracking | Data availability (0-1.0) per computation |

Requirements

• Python 3.9+
• Claude Code CLI
• httpx (HTTP client)
• fpdf2 (PDF generation)
• Optional: API keys for live AI platform data

License

MIT License

---

Built by JaeHyunGround | Powered by Claude Code