Kai Surround Sound

Kai root note: knowledge/, harness/, and scripts/ paths in this skill live in the Kai install, not the user's project. Resolve them against the first ancestor directory of this SKILL.md that contains a knowledge/ folder (the Kai plugin root, ~/.claude/kai, or the kai-cmo-harness repo). MARKETING.md, memory/, and any output files live in the current project. If a referenced scripts/ command is not available in this install, say so, skip it, and continue with the file-based guidance — never fabricate its output.

Build a consensus web that makes LLMs mention your brand. The bleeding edge of marketing.

Phase 0: Load Product Context

Check if MARKETING.md exists in the project root (same directory as CLAUDE.md, README.md, package.json).

If it exists: Read it — skip product discovery questions. It has the product name, ICP, value prop, monetization, brand voice, current channels, and competitive landscape.

If it does NOT exist: Auto-explore the codebase to create it in the project root (next to CLAUDE.md). Do NOT ask the user what the product is. Read CLAUDE.md, README.md, PROJECT.md, package.json, landing pages, and any project files. Search for email/ad/analytics config. Then create MARKETING.md using the template from /kai-email-system. Present draft to user for confirmation.

Phase 1: Entity Discovery

Read from MARKETING.md. Only ask about things not covered there:

1. Brand/product — what should AI mention?
2. Category — when someone asks "[what's the best X]", what's X?
3. Current AI presence — test: ask ChatGPT/Perplexity "what's the best [category]?" — are you mentioned?
4. Competitors mentioned — who IS being mentioned by AI when you're not?
5. Existing content assets — blog, social, press, directories, forums?

Phase 2: Consensus Audit

Load these before starting:

• knowledge/playbooks/surround-sound-llm-manipulation.md
• knowledge/frameworks/aeo-ai-search/llm-citation-tracking.md
• knowledge/frameworks/aeo-ai-search/entity-seo-knowledge-graph-deep-dive.md
• knowledge/frameworks/aeo-ai-search/perplexity-ranking-reverse-engineered.md
• knowledge/frameworks/aeo-ai-search/geo-academic-research-synthesis.md

Audit the Current Consensus Web

Map where your brand appears (and doesn't) across LLM training/retrieval sources:

Own-Domain Agent-Readiness Audit

Before you can surround-sound from third parties, your own site has to be legible to the agents routing back. If ChatGPT can't parse your homepage, every pulse you build elsewhere dead-ends.

Load: knowledge/checklists/agent-readiness-checklist.md

Run the checklist against the user's primary domain. Report:

Auto-run the linter if available:

python scripts/quality_gates/agent_readiness_lint.py https://<their-domain>

Any P0 failure blocks the rest of the plan until fixed — surround-sound spend on a site that agents can't read is wasted.

Prestige Pulse Score

Each mention type generates a "pulse" of authority:

• Forum post mentioning brand → 1 pulse
• Third-party review → 3 pulses
• "Best of" list inclusion → 5 pulses
• Wikipedia mention → 10 pulses
• Academic citation → 10 pulses

Estimate your total pulse score vs top 3 competitors.

Phase 3: Consensus Building Plan

Generate a 90-day plan to build your consensus web:

Month 0: Fix Own-Domain Agent-Readiness (P0 only)

Remediate every P0 failure from the agent-readiness audit before spending on outbound surround sound.

Skip to Month 1 only if the linter reports Pass on P0.

Month 1: Foundation (Entity Establishment)

Month 2: Amplification (Third-Party Presence)

Month 3: Dominance (Surround Sound)

Phase 4: Content Production for AI Visibility

AEO-Optimized Content Rules

Load: knowledge/frameworks/aeo-ai-search/aeo-ai-search-playbook-2026.md

For every content piece in the plan:

• Atomic facts — one verifiable claim per sentence
• Entity-first — name your brand explicitly, don't use pronouns early
• Structured for extraction — tables, lists, Q&A format that LLMs can parse
• Citation-worthy — include original data, research, or unique insights
• Information Gain — say something NOT already in the consensus (novelty signals)

Priority Content Types for AI Visibility

1. Comparison pages: "[Brand] vs [Competitor]" — LLMs love structured comparisons
2. Definition pages: "What is [your category]?" — entity establishment
3. How-to guides: "How to [solve problem your product solves]" — with your brand as the tool
4. Data/research: Original stats, surveys, benchmarks — citation magnets
5. FAQ pages: Structured Q&A that matches how people ask AI

Phase 5: Monitoring

Weekly AI Check

Test these queries in ChatGPT, Claude, and Perplexity:

• "What's the best [your category]?"
• "Compare [your brand] vs [competitor]"
• "[Your category] for [your ICP]"
• "How to [problem you solve]"

Track: Are you mentioned? What position? What's said about you? Is it accurate?

Phase 6: Output

workspace/surround-sound/
├── _consensus-audit.md          # Current state of your AI presence
├── _90-day-plan.md              # Month-by-month action plan
├── _content-production-queue.md # Content pieces to produce (feed to /kai-content-calendar)
├── _directory-checklist.md      # All directories to claim/optimize
├── _monitoring-queries.md       # Weekly AI check queries
└── _competitor-ai-presence.md   # What competitors look like in AI answers