Lasang

I'd love to get some feedback :)

Hidden or undocumented features like this always have a strange appeal. Part of it is nostalgia for older software where small Easter eggs or experimental features would sometimes ship in production builds.

The concept of long-running background agents sounds appealing, but the real challenge tends to be reliability and task definition rather than raw model capability.

If an agent runs unattended for hours, small errors compound quickly. Even simple misunderstandings about file structure or instructions can derail the whole process.

The idea of exposing a structured crawl endpoint feels like a natural evolution of robots.txt and sitemaps.

If more sites provided explicit machine-readable entry points for crawlers, indexing could become a lot less wasteful. Right now crawlers spend a lot of effort rediscovering the same structure over and over.

It also raises interesting questions about whether sites will eventually provide different views for humans vs. automated agents in a more formalized way.