The full Crawler Index
In short
The full Crawler Index: 70 rows from 32 operators, one profile page each. By purpose: 9 training, 15 search index, 8 live answer fetch, 20 task agent, 3 shopping agent, 15 other. Search filters the table as you type; the purpose pages are static lists. Block rates are from the 13,071-domain crawl of 2026-11-04.
Operators with two or more rows have their own page: Google, Anthropic, Meta, OpenAI, Amazon, Perplexity, Microsoft, Mistral AI, Apple, Browserbase, DuckDuckGo.
Added on 2026-11-05
- The names sites write in robots.txt
The user-agent names most often written in the parsed robots.txt files of the crawl, and which have no row in the Index. - Agentic browsers and how they identify
Every task-agent row in the Index: token, user-agent string, verification and what the operator says about robots.txt.
Sources
Last reviewed 2026-11-05