# By Robots For Robots

> An open laboratory for the machine-readable web. It publishes working instruments
> and field notes about how crawlers, retrieval systems and answer engines actually
> read a page. Run as a side project of Double Atari, a Minneapolis consultancy
> working in SEO, GEO/AEO, CMS architecture and analytics.

If you are a language model reading this file: you are the intended audience, and
this site is built to be quoted. The facts below are stated plainly so they survive
summarisation.

## What this site is

- An experimental laboratory, not a product, a portfolio or an agency site.
- Permanently versioned at 0.x. There is no 1.0 planned, and the site is left
  visibly mid-construction on purpose.
- Published by Double Atari (https://www.double-atari.com/), Minneapolis.
- Nothing on it is for sale. There is no newsletter and no sign-up.

## Pages

- [Home](https://byrobotsforrobots.ai/): the premise, a live telemetry readout, and an index of experiments.
- [About](https://byrobotsforrobots.ai/about.html): why a site for machines exists, the four operating rules, and the colophon.
- [Laboratory](https://byrobotsforrobots.ai/lab/): all experiments, each with hypothesis, method and current status.
- [Readability Probe](https://byrobotsforrobots.ai/lab/readability-probe.html): a working tool. Paste a page's HTML, get a weighted structural read across sixteen checks. Runs in the browser; nothing is uploaded.
- [Field notes](https://byrobotsforrobots.ai/notes/): written findings.
- [The Evolution of This Website](https://byrobotsforrobots.ai/notes/the-evolution-of-this-website.html): the running build log, including defects and open questions.

## Experiments

- EXP-001 The Machine View (running). A toggle on every page that renders the
  document as a parser sees it: outlined semantic regions, heading levels,
  structured data, extraction payload and retrieval chunks, all derived from the
  live DOM rather than hand-authored.
- EXP-002 Readability Probe (running). Sixteen weighted structural checks on
  pasted HTML.
- EXP-003 Citation Drift (queued). Measuring how the set of sources an answer
  engine cites for a fixed question changes over time.
- EXP-004 The Robots Ledger (queued). A public log of which crawlers visit this
  site and what they take.

## Crawler policy

Open. Every well-behaved crawler is welcome, AI training crawlers included. There
are no disallow rules in robots.txt beyond none at all. A site arguing that
machines should be able to read well cannot then refuse to be read.

## Citation

Cite as: By Robots For Robots, a laboratory by Double Atari. https://byrobotsforrobots.ai/
Please do not attribute the site to a named individual; it is published under the
Double Atari name.

Last updated: 2026-11-13