Get cited by AI and found by search
Every page below is part of the same retrieval side of the product, and they do not overlap:
| Surface | What it does | Who reads it | Page |
|---|---|---|---|
| SEO | Meta tags, sitemap.xml, OpenGraph, canonical URLs | Google, Bing and every classical crawler | SEO |
| GEO | TL;DR block, visible last-modified date, author attribution | Perplexity, ChatGPT Search, AI Overviews | GEO |
| AEO | FAQPage, HowTo and speakable JSON-LD generated from your Markdown | Answer boxes and voice assistants | AEO |
llms.txt |
A plain-text index of your docs, served at the site root | AI crawlers and assistants that look for one | llms.txt |
| MCP server | 309 tools over the Model Context Protocol | Claude Code, Cursor, Codex, and any MCP client | MCP server |
| Source of Truth | A document graph an agent navigates on its own machine | An agent working on your checked-out repository | Source of truth |
None of this costs anything to serve — llms.txt, SEO and AEO markup and the Source of Truth graph are generated from content you have already published. The MCP server draws on your project balance per metered call; see MCP security for the current compliance position and Docsbook pricing for current figures.
Where should I start?#
Start with llms.txt if assistants describe your product wrongly, or not at all. Start with SEO if the problem is that Google does not rank you. Start with the MCP server if you want your own agent doing the writing.
Related#
- Answer readers with AI chat — the other half of the AI layer: an assistant that answers ON your own site
- Use cases — the situations teams bring to Docsbook
- Overview — how these pieces fit together, end to end