Skip to content

Appendix

This is the book's reference layer: the pages you open while doing the work, not while learning it. Everything here is distilled from the chapters — the chapters carry the reasoning and the failure modes; the appendix carries the lookup tables and the copy-paste starting points. If a template or table here seems to contradict a chapter, the chapter wins (and please file an issue).

What's in the appendix

Page What it is Reach for it when…
Tool directory Every external tool the book references — Google's consoles and APIs, Bing, schema validators, DNS/email diagnostics, crawl probes, free traffic-estimation sources, MCP directories — each with what it's for, a URL, and a free/paid note You know what to check but not where
Skill index The chapter → source-skill map into the public ever-just/agentskills repo You want the operational, agent-runnable version of a chapter's method
Templates Eight copy-paste starters: LocalBusiness @graph, SaaS dual-node @graph, FAQPage, AI-crawler robots.txt, llms.txt, MCP server.json, policy-safe review markup, the Ask-AI deep-link row You're shipping today and need a validated skeleton, not a tutorial
AI crawler registry The user-agent table: operator, what each bot feeds (training / search index / live fetch), robots.txt behavior, verification method, robots token — plus the curl-as-bot test pattern You're writing robots rules, reading server logs, or debugging a bot wall

How to use it

  1. Templates are skeletons, not answers. Every value is a placeholder (example.com, YOUR-NAME). Replace all of them, then validate — each template's usage note names the validator.
  2. Volatile facts are date-stamped. Crawler behavior, deep-link URL schemes, and API scopes change; anything marked "as of 2026" should be re-verified before you build automation on it (the book re-verifies quarterly).
  3. Evidence tiers are explicit. The crawler registry distinguishes operator-documented behavior from community-reported behavior. Never treat the second kind as a guarantee.
  4. The skills repo is upstream. New frameworks land in ever-just/agentskills first, then get distilled into chapters here — the skill index explains the loop.

The fastest paths through