llms.txt Is the Cheapest RAG You'll Ever Ship
My site’s chat agent answers with exact deep links to my posts - no search tool, no vector database. It reads /llms.txt at prompt time, and the whole ‘retrieval pipeline’ is one HTTP GET.
My site’s chat agent answers with exact deep links to my posts - no search tool, no vector database. It reads /llms.txt at prompt time, and the whole ‘retrieval pipeline’ is one HTTP GET.
Wire an xpander agent’s success/error webhook into hark.ryan.ceo and your phone buzzes the moment a long-running agent task finishes - no polling, no forgotten tabs.
The “David’s AI” sidebar is a real xpander agent - created with one API call, streamed with another, and dropped onto a static site behind a tiny key-holding Worker.
Instead of full delegation, we are literally working harder, not smarter.
AI agents only create durable productivity when they remove humans from being the integration layer between systems.
It might be the most useful AI layer: watching the mess, extracting the work, and telling you what needs to happen next.
The gap between “we have 5 agents” and “we have 500 agents” is not filled by buying more products. It is filled by building the foundation.
An external vertical agent is not an internal agent. A platform that wraps your code and takes over the runtime is not really infrastructure.
How AI agents change the traditional API development model: skip the server layer and let agents work directly with API contracts.
Upload an invoice and the agent reads it, classifies it, renames it, files it into the right Google Drive folder, and keeps an audit trail: the parts that break in production, solved.