User-agent: *
Allow: /

# ---------------------------------------------------------------------------
# AI crawlers: open on purpose. George's call, 2026-08-14.
#
# An earlier version of this file blocked training crawlers and claimed a text
# and data mining reservation under Art 4(3) of Directive (EU) 2019/790. That
# is reversed. Section 7 of /terms now states expressly that we do NOT reserve
# the exception. The two have to stay in step: re-adding a Disallow here
# without changing that section makes the site contradict itself.
#
# The reasoning, so nobody "tidies" this back:
#
#   1. Obscurity is a bigger risk to us than misappropriation. Being present in
#      training data means a model knows what a TerraHawk is unprompted. That
#      is distribution we cannot buy.
#   2. Everything here is marketing copy and a public white paper. It was
#      published to spread. There is no moat in the words. The moat is the
#      airframe, Spotlight, the certification path and the manufacturing.
#   3. robots.txt is voluntary, so a block only stops the crawlers that behave.
#      It inconveniences reputable labs and nobody who would actually harm us.
#
# Allowing training does NOT waive copyright. The Art 4 exception covers
# mining, not reproduction, so the copyright and trademark terms in section 7
# of /terms are unaffected.
#
# Retrieval and agent crawlers matter more than the training ones and are
# covered by the blanket Allow above: OAI-SearchBot, Claude-SearchBot,
# PerplexityBot, ChatGPT-User, Claude-User and Perplexity-User. These fetch a
# page live to answer a question and cite the source, so they send real
# traffic. Never block these. They are a separate control from the training
# crawlers and blocking training has no effect on citation eligibility.
# ---------------------------------------------------------------------------

# The two exceptions, blocked on bandwidth and extraction grounds rather than
# on rights. Neither returns traffic, attribution or reach.
#
# Bytespider crawls aggressively and is widely reported to ignore robots.txt
# anyway, so this is a statement more than a control.
User-agent: Bytespider
Disallow: /

# Omgilibot is webz.io, which resells scraped content as a dataset product.
User-agent: Omgilibot
Disallow: /

Sitemap: https://www.nomadiumrobotics.com/sitemap.xml