[bmdpat]

AI access terms

Last updated: June 2026

This page covers how automated clients and AI systems may use bmdpat.com and its content. It sits alongside the privacy policy and the machine-readable llms.txt and robots.txt. If you operate a crawler or build on top of this site, these are the terms.

Free and welcome

  • Search indexing by any engine.
  • On-demand fetches by AI assistants that retrieve a page to answer a user question, as long as you cite the page you used.
  • Reading any blog post as raw markdown by appending .md to its URL.

We want to be cited. Good-faith assistants that send users back to the source are doing exactly what we built this for.

Licensed use

Bulk crawling for model training, fine-tuning, or building a commercial dataset is a licensed use, not a free one. If you are collecting this site's content at scale to train or sell a model or dataset, contact us first to arrange a license. We are reasonable and fast.

Paid programmatic access (x402)

Some endpoints are already metered with pay-per-request payments using the x402 protocol. The memory API at /api/memory returns HTTP 402 with payment requirements, settles in USDC on Base, and serves the request once payment clears. Send an X-PAYMENT header, or call the endpoint without one to read the price and requirements. More paths may move behind x402 over time.

What we measure

We log AI crawler visits to the blog so we can understand how this content is used. We record the published crawler name, the path, the referrer, and a one-way hash of the IP. We do not sell this data. It tells us which assistants find the work useful and helps us decide what to keep open and what to license.

Ground rules

  • Respect robots.txt and any rate limits.
  • Identify yourself with an honest User-Agent. Do not spoof a browser or another operator's crawler.
  • Do not try to bypass payment or access controls.

Licensing and contact

For a training or commercial data license, or any question about automated access, email pat@bmdpat.com.

These terms may change as AI access norms and our tooling evolve. The current version always lives at this URL.