RSS feeds supposed to be the cheapest way to read data from your website, they are not pulling whole 10MB of javascript and other assets, but, yeah, let's block those. Everyone will just send their "llm agents" to pull data from regular website anyway.

The assets aren't a problem in 2026, they are in an edge cache.

The RSS feed can't be cached for long, and AI crawlers will be making millions of requests to these.