# AI SEO

> Use when optimizing a site or brand to show up in AI answers, LLM citations, AI search, product recommendations, and category answers.

- Skill: `infinite-labs-ai/ai-seo` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add infinite-labs-ai/ai-seo`
- Raw SKILL.md: https://api.skillmd.com/api/skills/infinite-labs-ai/ai-seo/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Marketing & Growth
- Author: Infinite-Labs-AI (https://skillmd.com/u/infinite-labs-ai)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/infinite-labs-ai/ai-seo

---


# AI SEO

Make the company, product, category, use cases, and proof easy for AI answer engines to understand and cite.

## Audit The Entity

Inspect what the web can verify:

- Official site pages that explain the product and category.
- Founder, team, or company pages.
- Docs, pricing, comparisons, case studies, changelog, integrations.
- Third-party mentions, directories, reviews, podcasts, GitHub, app stores, communities.
- Structured data and clean metadata.

If live research is needed, use current sources and prefer primary pages.

## Diagnose Answerability

Ask:

- Can an AI system identify what this is in one sentence?
- Is the category explicit?
- Are use cases and target customers named?
- Are claims supported by public proof?
- Are comparisons stated cleanly without legal or factual overreach?
- Are there third-party corroboration points?
- Can important pages be crawled without app login or script-only rendering?

## Improve The Evidence

Inventory public evidence by claim:

- Claim the company wants answer engines to repeat.
- Official source that supports it.
- Third-party corroboration.
- Query or buying question it helps answer.
- Evidence strength.
- Missing proof or crawl problem.

Then recommend assets that make weak claims easier to verify: canonical explainer, use case page, public example, customer story, fair comparison, documentation, review surface, or reputable third-party mention.

## Guardrails

- Do not recommend fake reviews, synthetic citations, or spam directories.
- Do not overstate unsupported claims.
- Treat answer engines as retrieval and synthesis systems, not magic keyword boxes.
- Prefer durable public evidence over prompt-hack phrasing.

## Output

```text
AI answer presence read:
[one paragraph]

Entity clarity:
- Category:
- Audience:
- Use cases:
- Proof:
- Third-party corroboration:

Gaps:
1. [gap] -> [why it matters]

Evidence table:
| Claim | Official source | Third-party source | Query supported | Strength | Gap |

Assets to create:
| Asset | Query/job it supports | Proof required | Priority |

Schema and crawl fixes:
- [fix]

30-day presence plan:
1. [action]
2. [action]
3. [action]
```

