AI Crawler Atlas / domains
github.blog: AI crawler policy in robots.txt
robots.txt as fetched on 2026-10-10: github.blog does not disallow any of the 22 AI and search crawlers we check from the site root.
The file asks nothing of these crawlers, so it is open to AI training, AI answers and live assistant fetches as far as robots.txt goes.
This reports what the file asks. It does not show whether a crawler can actually reach the site: many sites also block crawlers at their CDN or firewall, and some crawlers ignore robots.txt.
Training crawlers
Disallowing these keeps a site's content out of future model training. It does not remove what was already collected.
| Crawler | Operator | Status | How the file treats it |
|---|---|---|---|
| GPTBot | OpenAI | not disallowed | covered by the default rule |
| ClaudeBot | Anthropic | not disallowed | covered by the default rule |
| Google-Extended | not disallowed | covered by the default rule | |
| CCBot | Common Crawl | not disallowed | covered by the default rule |
| meta-externalagent | Meta | not disallowed | covered by the default rule |
| Bytespider | ByteDance | not disallowed | covered by the default rule |
| Applebot-Extended | Apple | not disallowed | covered by the default rule |
| Amazonbot | Amazon | not disallowed | covered by the default rule |
| cohere-ai | Cohere | not disallowed | covered by the default rule |
| Diffbot | Diffbot | not disallowed | covered by the default rule |
| omgili | Webz.io | not disallowed | covered by the default rule |
| Timpibot | Timpi | not disallowed | covered by the default rule |
Search and citation crawlers
These feed AI answer engines that cite and link their sources. Disallowing them can take a site out of AI answers and out of the referral traffic that comes with them.
| Crawler | Operator | Status | How the file treats it |
|---|---|---|---|
| OAI-SearchBot | OpenAI | not disallowed | covered by the default rule |
| PerplexityBot | Perplexity | not disallowed | covered by the default rule |
| Claude-SearchBot | Anthropic | not disallowed | covered by the default rule |
| DuckAssistBot | DuckDuckGo | not disallowed | covered by the default rule |
User-action fetchers
These fetch a page live when a person asks an assistant about it. Disallowing them stops the assistant reading the page for that person.
| Crawler | Operator | Status | How the file treats it |
|---|---|---|---|
| ChatGPT-User | OpenAI | not disallowed | covered by the default rule |
| Perplexity-User | Perplexity | not disallowed | covered by the default rule |
| Claude-User | Anthropic | not disallowed | covered by the default rule |
| meta-externalfetcher | Meta | not disallowed | covered by the default rule |
Classic search crawlers
Classic search crawlers, shown as the baseline.
| Crawler | Operator | Status | How the file treats it |
|---|---|---|---|
| Googlebot | not disallowed | covered by the default rule | |
| Bingbot | Microsoft | not disallowed | covered by the default rule |
How github.blog compares
Among 7,107 readable sites in the Top 10,000, 16.6% disallow GPTBot, 15.2% ClaudeBot, 7.3% OAI-SearchBot and 11.2% PerplexityBot.
History
First read 2026-10-10; the file has not changed since. Read the file itself.
Check it live. This page shows the file as read on 2026-10-10. Run a live check of github.blog and get told when it changes, or get the dataset for a segment.