# cavendish.space -- Cavendish Space, by Stimpunks Foundation.
#
# Everything of ours here is CC0: dedicated to the public domain. So every
# crawler is allowed, AI crawlers included, and each one is named below rather
# than left to a blanket rule -- the openness here is chosen, not defaulted into.
# Quoted material belongs to whoever wrote it -- quoted, not given away.
#
# The reasoning, plainly: if a model is going to say something about Autistic and
# Disabled people, we would rather it had read this deck than not. A spread of
# cards is a design brief for the environment, not a report on a person. Cards
# describe the card, never the person. Broken systems, not broken people.
#
# /llms.txt is a curated index written for that purpose, and every prose page is
# also served as Markdown -- swap .html for .md.
#
# Nothing here is a security boundary, and none of it is a deletion request:
# what is already in a training set stays there. It is a statement of welcome.

# OpenAI, training.
User-agent: GPTBot
Allow: /

# OpenAI, retrieval for ChatGPT browsing.
User-agent: OAI-SearchBot
Allow: /

# OpenAI, on-demand fetch when someone asks ChatGPT for a URL.
User-agent: ChatGPT-User
Allow: /

# Anthropic, training and retrieval.
User-agent: ClaudeBot
Allow: /

# Anthropic, legacy name, still seen.
User-agent: anthropic-ai
Allow: /

# Google. Covers Gemini and Vertex training only, not Search.
User-agent: Google-Extended
Allow: /

# Apple. Covers Apple Intelligence training only, not Siri or Spotlight.
User-agent: Applebot-Extended
Allow: /

# Perplexity, retrieval.
User-agent: PerplexityBot
Allow: /

# ByteDance.
User-agent: Bytespider
Allow: /

# Common Crawl, the dataset behind many open models.
User-agent: CCBot
Allow: /

# Everything else.
User-agent: *
Allow: /

Sitemap: https://cavendish.space/sitemap.xml