AI agents are about to become software buyers. Is anyone else thinking about this?

I've been digging into how AI agents interact with SaaS products, and I think there's a gap that hasn't been discussed much yet. When an agent tries to evaluate or use a SaaS tool for a user, it essentially has to scrape your marketing page like it’s 2009. There’s no standard way to find pricing, understand what the product actually does, or complete a purchase without going through a human-controlled checkout process that disrupts everything. Three solutions partially address this issue: llms.txt - A plain text file at your domain root that informs agents of your pricing, policies, and capabilities. It’s like robots.txt, but for LLMs. The spec exists, but few have adopted it. MCP servers - These allow you to expose your product's core actions as callable tools, enabling an agent to invoke functions like list_plans() or create_project() directly. The spec is available, but most SaaS products haven't used it. Agent checkout protocols - These include systems like ACP that enable an agent to complete a purchase without redirect flows or confirmation screens that assume a human is overseeing the process. What keeps bothering me is that the conversion of human visitors is already shifting as more research and decision-making gets passed to agents. If your product can't be found or evaluated by a non-human, you could be missing out on deals without even realizing it. Has anyone noticed agent traffic in their analytics? Have you intentionally implemented any of these three solutions, or are they still off the radar? Would you consider paying for a solution that manages this layer for you, or is this something you’d prefer to handle in-house?