Content Intent Agent

Content Intent Agent Skill

Purpose

Verifies every piece of content matches user search intent to maximize:

• Sales (e-commerce, products, services)
• Leads (quotes, consultations, demos)
• Digital sales (courses, memberships, digital products)
• Mailing list growth (email subscribers, newsletter signups)

34 verification modules + 16-part scoring system = Complete conversion optimization framework with veto power.

Prerequisites

Requires .claude-project-config.json in project root with:

• Jurisdiction & compliance frameworks
• Target demographics (age, income, disabilities, location)
• Psychographics (pain points, desires, fears, buying triggers)
• Conversion goals (primary and secondary)
• Intent rules (query classification, required modules)
• Offer library (lead magnets, calculators, coupons)
• KPI requirements (analytics events to track)
• Conversion optimization settings

User must research and populate all sections before using skill.

THE 34 VERIFICATION MODULES

TRADITIONAL SEARCH INTENT (1-15)

1. Deep Semantic Intent Matching

Verifies headlines/CTAs mirror search query + intent modifiers ("buy", "how", "reviews").

Veto if: H1 missing primary keyword, intro doesn't address query in first 100 words.

2. Query Classifier & Stage Inference

Auto-classifies: transactional/commercial/informational/navigational/local/branded/problem-aware Maps to: BOFU/MOFU/TOFU

Veto if: TOFU content with hard "Buy Now" CTA, BOFU with weak "Learn More".

3. Intent-Template Validator

Enforces required components per intent:

• Transactional: above-fold price/CTA, trust badges, phone tap, financing, FAQ, returns
• Commercial: comparison table, pros/cons, "best for" segments, social proof, soft CTA
• Local: NAP, service area map, call button, hours, GBP cues
• Informational: answer box, step list, FAQ schema, lead magnet

Veto if: Missing any required module for detected intent class.

4. Dynamic User Intent Feedback Loop

Simulates persona flows - validates target users can complete desired actions.

Veto if: Persona simulation shows confusion, can't complete conversion, trust insufficient.

5. Competitor Intent Gap Analysis

Analyzes top 10 SERP competitors for same keywords, flags missing USPs/trust elements.

Veto if: Missing critical elements present on all top 3 competitors.

6. SERP Pattern Mirror & Differentiation

Parses top-10 patterns (lists, tables, FAQs, calculators), enforces common modules + 1 unique element.

Veto if: Missing patterns present in 70%+ of top 10, or zero differentiation.

7. Ad-to-Landing Parity Guard

Validates headline/USP/price/phone consistency between ads and landing pages.

Veto if: Ad keyword not in H1, price differs, phone number differs, USP absent.

8. Negative-Intent & Disqualifier Filter

Detects mismatched sub-intents ("free", "DIY", "jobs"), forces alternate offers.

Veto if: Hard conversion CTA on "free" intent, no lead magnet on informational query.

9. Offer/Lead Magnet Selector

Maps offer type to intent + persona: pricing PDF for BOFU B2B, coupon for ecommerce.

Veto if: Wrong offer type for intent, no offer for MOFU/TOFU traffic.

10. Trust Signal Orchestrator

Manages persona-weighted placement of logos, testimonials, certifications, guarantees.

Veto if: No trust signals above fold BOFU, outdated proof, trust density insufficient for high-ticket.

11. Conversion Readiness Scorecard

Computes 16 scores (0-100 each): Intent Fit, Scent, Trust, UX Friction, Accessibility, Offer Fit, LLM Optimisation, Affiliate Performance, Momentum Flow, Objection Coverage, Urgency Integrity, Exit-Rescue, Revenue Expansion, Multi-Device, CRM Tagging, Zero-Friction.

Veto thresholds: See config - typically 75-100 per score.

12. Schema-by-Intent Enforcer

Validates structured data: Product/Service, FAQ, Review, LocalBusiness, HowTo, ItemList, Breadcrumb.

Veto if: Missing required schema for intent class.

13. Calculator/Quote Trigger

Detects pricing intent ("cost", "price", "calculator"), enforces working estimator.

Veto if: Pricing intent without calculator/quote form, or non-functional.

14. Language/Locale & Variants Guard

Enforces jurisdiction-specific: spelling (US/UK/AU/CA), currency, units, date/phone formats.

Veto if: Wrong spelling variant, wrong currency, wrong measurement units, generic examples.

15. Cannibalisation & Routing Check

Detects multiple pages targeting same intent/keyword, suggests consolidation.

Veto if: New page duplicates existing page intent without clear differentiation.

COMPLIANCE & TRACKING (16-17)

16. Compliance & Claims Gate

Jurisdiction-aware enforcement:

• US (FTC): Material connection disclosure, testimonials reflect typical results, health claims substantiated, up to $43,792 per violation
• Australia (ACL): Testimonials genuine/verifiable, "results not typical" required, up to $1.1M fines
• UK (ASA): Claims substantiated pre-publication, comparative claims need proof
• EU (GDPR): Privacy policy required, right to be forgotten, environmental claims substantiated
• Canada: Performance claims need testing, "regular price" must be genuine

Veto if: Superlative without proof, fake testimonials, medical claims without disclaimers, missing compliance for jurisdiction.

17. KPI Instrumentation Hooks

Asserts events: view_item, generate_lead, begin_checkout, phone_click, scroll_depth, form_error.

Veto if: Missing required events, phone not clickable (tel: link), UTM parameters malformed.

ADVANCED CONVERSION (18-21)

18. Golden Path & Fallback CTA

Primary CTA + friction-reduced fallback (phone, chat, "email me pricing", WhatsApp).

Veto if: Only one CTA, fallback not truly reduced friction, no low-commitment option TOFU/MOFU.

19. Automated UX & Accessibility Scans

Scans forms/CTAs for: touch targets (44-48px), contrast (WCAG AAA for elderly), labels, mobile responsiveness.

Veto if: Touch targets <44px (elderly demographic), contrast below WCAG AAA, form missing labels, mobile broken.

20. AI-Powered Content Personalisation

Proposes segment-specific adjustments: headline variations for elderly vs B2B, CTA rotation for new vs repeat.

Veto if: No personalisation strategy for multi-persona pages, single message for conflicting personas.

21. Intent-Driven A/B Testing Protocol

Mandatory split-testing: headline, CTA button, trust elements per intent.

Veto if: No testing plan for new high-value pages, can't measure results.

LLM/AI SEARCH (22)

22. LLM/Conversational Search Intent

Optimizes for ChatGPT, Claude, Perplexity, SearchGPT, Google SGE, voice assistants.

Components:

• Conversational Intent Classifier (question type, user stage, pain level)
• Answer Completeness Validator (primary + all sub-questions answered)
• Context Awareness Checker (acknowledges user's situation)
• Follow-up Question Anticipator (pre-answers obvious next questions)
• Citation/Source Optimisation (complete sentences for LLM extraction)
• Conversational Tone Matcher (matches tone to query emotion)
• Multi-Modal Answer Structure (voice, visual AI, code extraction)
• Natural Language Query Mapping (tracks queries to content sections)
• AI Search Behaviour Adaptation (platform-specific preferences)

Veto if: Primary question answered <150 words (too shallow), obvious sub-questions not addressed, key facts in bullet fragments LLMs can't extract.

AFFILIATE CONTENT (23)

23. Affiliate Offer Intent Matching

A. Offer-Intent Alignment: Recommended product genuinely solves stated problem, budget appropriate, commission vs quality resolved in user's favor.

B. Recommendation Credibility: Clear methodology, specific testing details, cons disclosed, update dates visible.

C. Affiliate Disclosure Compliance (Multi-Jurisdiction):

• US (FTC): "We earn from qualifying purchases" - clear & conspicuous, $43,792 per violation
• Australia (ACL): Plain English before links, $1.1M fines
• UK (ASA): "#ad" or "Advertisement" labels
• EU: "Advertising" or "Sponsored" required
• Canada: Material connection per recommendation

D. Comparison Fairness: Objective criteria, all relevant products, better non-affiliate options included.

E. Multi-Merchant Conversion: Multiple retailers, price comparison, in-stock status.

F. Trust Architecture: Editorial independence, expert credentials, testing transparency.

Veto if: Product doesn't meet requirements, budget ignored, no disclosure, disclosure non-compliant, comparison biased.

CONVERSION ENGINEERING (24-34)

24. Conversion Momentum Sequencing

Enforces micro-commitments: quiz → results → email → consultation (not giant leap).

Impact: +35-60% conversion. Veto if: Jumps straight to hard conversion, no micro-options, dead ends.

25. Intent-Matched Objection Handling

Pre-answers objections at decision points:

• BOFU pricing: "Why this price?" below price
• Local: "What if not in my area?" above service map
• Elderly: "Is this safe/easy?" near CTA
• B2B: "What's ROI?" before pricing

Impact: +20-35% completion. Veto if: Obvious objections not addressed, objections buried.

26. Post-Search Proof Density Guard

Zone-based minimums:

• Hero: 1 trust signal above fold
• Mid-page: 2-3 proof elements
• CTA zone: 2 within 300px (3-4 for high-ticket)

CRITICAL - FAKE TESTIMONIAL DETECTION (ZERO TOLERANCE):

✅ Required:

• Real people: Full name, city, photo with permission
• Verifiable: Can provide proof if challenged
• Current: Within 18 months or marked [year]
• Disclosure: If paid/incentivized, must disclose

❌ INSTANT VETO:

• Stock photos
• AI-generated reviews
• Generic names ("John S.")
• Unverifiable claims
• Exceptional results without disclaimer

Consequences: FTC $43,792 per fake, ACL $1.1M, reputation destroyed.

Impact: +15-40% lift. Veto if: Zero proof above fold, ANY fake testimonials = INSTANT VETO.

27. Ethical Urgency & Scarcity Automation

✅ Allowed: Real inventory, actual calendar, genuine promos, true batch limits.

❌ Blocked (INSTANT VETO): Countdown resets, "Only X left" with infinite stock, fake counters, manufactured urgency.

Impact: +30-80% conversion. Veto if: ANY fake scarcity = INSTANT VETO.

28. Lead-Rescue Automation

Exit intent + idle triggers with intent-matched offers.

Impact: +20-50% additional capture. Veto if: No exit capture, intrusive, triggers repeatedly.

29. Revenue Expansion Layer

Upsells/cross-sells at decision point.

Impact: +15-40% AOV. Veto if: No expansion options, poorly positioned, forced bundles.

30. Multi-Device CTA Continuity

Hand-offs: Mobile→Desktop, Desktop→Mobile, cross-session.

Impact: +20-35% from multi-device users. Veto if: No hand-off, cart doesn't persist.

31. Search-Intent CRM Tagging

Tags leads: intent_type, funnel_stage, query, device, engagement.

Impact: +25-60% email→sale conversion. Veto if: Leads without intent tags, follow-up not differentiated.

32. Zero-Friction Conversion Paths

Express checkout, guest checkout, pre-filled forms, social login, max 5 fields.

Impact: -30-50% abandonment, +40-60% mobile conversion. Veto if: No express checkout, account required, >5 required fields.

33. Exit Intent Precision Rescue

Behavior-specific (not generic): viewed pricing → discount, used calculator → save results.

Impact: 20-40% of abandoning visitors captured. Veto if: Generic popup, triggers <10 seconds, covers content.

34. Live Engagement & AI Agent

Real-time assistance at high-value moments: pricing, comparison, checkout.

Impact: +20-35% for high-ticket/B2B. Veto if: No live option for high-ticket (>$2,000), chat missing on pricing.

16-PART CONVERSION READINESS SCORECARD

All scores 0-100:

1. Intent Fit (< 85 = veto) - H1/intro/body/CTA alignment
2. Scent (< 80 = veto) - Ad↔landing parity
3. Trust/E-E-A-T (< 70 = veto BOFU) - Proof variety/recency
4. UX Friction (< 75 = veto) - Form fields, mobile, load
5. Accessibility (< 80 = veto elderly) - Contrast, labels
6. Offer Fit (< 75 = veto) - Lead magnet appropriate
7. LLM Optimisation (< 75 = veto conversational) - Answer completeness
8. Affiliate Performance (< 75 = veto affiliate) - Credibility, disclosure
9. Momentum Flow (< 85 = veto) - Micro-commitments present
10. Objection Coverage (< 80 = veto) - Top 3 objections addressed
11. Urgency Integrity (100 required) - NO fake scarcity
12. Exit-Rescue (< 75 = veto) - Exit capture present
13. Revenue Expansion (< 70 = veto) - Upsell opportunities
14. Multi-Device Continuity (< 75 = veto) - Hand-offs exist
15. CRM Intent-Tagging (100 required) - All leads tagged
16. Zero-Friction (< 75 = veto) - Express checkout, min fields

CONFIG FILE SCHEMA

Create .claude-project-config.json in project root:

{
  "project_name": "Your Project",
  "jurisdiction": {
    "primary_country": "AU",
    "compliance_frameworks": ["ACL"],
    "language_locale": "en-AU",
    "currency": "AUD",
    "measurement_units": "metric"
  },
  "target_demographics": {
    "age": "65-85",
    "income": "middle to upper middle",
    "disabilities": "mobility impaired",
    "location": "Australia"
  },
  "psychographics": {
    "pain_points": ["loss of independence", "fear of falls"],
    "desires": ["maintain dignity", "stay active"],
    "buying_triggers": ["doctor recommendation"]
  },
  "conversion_goals": {
    "primary": "phone_enquiries",
    "secondary": ["email_leads"]
  },
  "intent_rules": {
    "classes": ["transactional", "commercial", "informational", "local"],
    "required_modules": {
      "transactional": ["price_block", "primary_cta", "trust_cluster"]
    },
    "schema_required": {
      "transactional": ["Product", "FAQ"]
    }
  },
  "offer_library": [
    {"id": "buyers_guide", "fits": ["informational"], "goal": "email"}
  ],
  "kpi_assertions": {
    "required_events": ["generate_lead", "phone_click"],
    "utm_policy": "required"
  },
  "psychological_triggers": {
    "elderly": ["safety", "independence", "dignity"]
  },
  "conversion_optimization": {
    "momentum_sequencing": {"enabled": true},
    "urgency_rules": {"fake_scarcity_blocked": true},
    "exit_rescue": {"enabled": true},
    "zero_friction": {"express_checkout_required": true}
  },
  "veto_thresholds": {
    "intent_fit": 85,
    "urgency_integrity": 100,
    "crm_intent_tagging": 100
  }
}

USAGE

Activation

PM orchestrator directs agent when verification needed:

Create landing page for "buy mobility scooter Brisbane"
then verify with content-intent-agent skill

Agent:

1. Reads .claude-project-config.json from project root
2. Evaluates content against all 34 modules
3. Computes 16-part scorecard
4. Returns APPROVED or VETO with detailed feedback

Feedback Format

VETO - Critical issues prevent shipping

Intent Fit: 72 (threshold 85) ❌
Momentum Flow: 68 (threshold 85) ❌

CRITICAL ISSUES:

1. Line 12, H1: Missing transactional intent modifiers
   Why: Informational framing on BOFU query depresses conversion 50-70%
   Fix: Change to "Buy Kymco Mini Comfort - Price & Free Delivery"
   
2. Module 26: Stock photo testimonial detected
   Why: FTC violation $43,792, destroys trust
   Fix: Remove stock photo, use real customer with signed release

EXPECTED RESULTS

With all 34 modules:

• Overall conversion: +60-120%
• Lead capture: +50-80%
• Phone calls: +40-60%
• AOV: +15-40%
• Mobile conversion: +40-60%
• Email→sale: +25-60%
• CPA: -30-50%

CRITICAL COMPLIANCE

ZERO TOLERANCE:

1. Fake testimonials (stock photos, AI, unverifiable) = FTC $43,792 per, ACL $1.1M
2. Fake scarcity (countdown resets, false "only X left") = Regulatory violations
3. Deceptive claims (superlatives without proof) = Fines + reputation damage

VERSION

v1.3.0 - Complete with 34 modules + 16-part scoring

SUMMARY

Complete conversion optimization system verifying content against 34 criteria to ensure maximum sales, leads, and list growth.

Key principles:

1. User intent first
2. Conversion engineering
3. Ethical compliance (zero fake tactics)
4. Multi-platform (traditional + AI search)
5. Data-driven (16-part scorecard)

Every visitor has a job to be done. This skill ensures content helps them do that job while achieving business goals.