# robots.txt for vtex.com
# Last updated: 2026-10-28

# Default: allow all crawlers
User-agent: *
Allow: /

# Disallow sensitive and duplicate paths
Disallow: /api/
# Social preview images are public, deterministic, and safe for unfurl crawlers.
Allow: /api/social-card
Disallow: /sign-in
Disallow: /sign-up
Disallow: /404
Disallow: /documents/*/legal/
Disallow: /documents/en-us/ParadigmB2B_Combine-26_vendor-Enterprise_VTEX.pdf

# Block cookies policy PDF from indexing (redirected to compliance.vtex.com)
Disallow: /*/trust/cookies-information/
Disallow: /*/trust/platform/cookies-information/

# Sitemaps
Sitemap: https://www.vtex.com/sitemap.xml
Sitemap: https://www.vtex.com/en-us/sitemap.xml
Sitemap: https://www.vtex.com/pt-br/sitemap.xml
Sitemap: https://www.vtex.com/es-mx/sitemap.xml

# ─── AI crawlers ─────────────────────────────────────────────────────────────
# One page, and one page only: the Cybernews statement. It already carries a
# noindex, nofollow directive (metadataConfig) as an HTTP header AND as a meta
# tag in the head, which keeps it out of Google and Bing. That does nothing
# about AI crawlers, which do not index — they ingest — so the directive they
# read is Disallow, and until 2026-09-12 ai.txt handed them a blanket
# "Allow: /en-us/" that explicitly covered this page.
#
# THIS BLOCK REPEATS THE GLOBAL RULES ON PURPOSE. In robots.txt a named
# user-agent group REPLACES the "*" group for that crawler — it does not
# inherit from it. Listing only the Disallow below would have silently handed
# these bots /api/, /sign-in and the rest.
#
# AND THE "*" GROUP DELIBERATELY DOES NOT GET THIS ENTRY. A Disallow stops the
# crawl, so Googlebot would never fetch the page and never read the noindex
# that de-indexes it — a blocked URL can still surface as a bare link. Disallow
# and noindex are alternatives, not layers; see the note above this function.
# Each is aimed at the crawler it actually works on.
User-agent: GPTBot
User-agent: OAI-SearchBot
User-agent: ChatGPT-User
User-agent: Google-Extended
User-agent: CCBot
User-agent: anthropic-ai
User-agent: ClaudeBot
User-agent: Claude-User
User-agent: Claude-SearchBot
User-agent: PerplexityBot
User-agent: Perplexity-User
User-agent: Bytespider
User-agent: cohere-ai
User-agent: Applebot-Extended
User-agent: meta-externalagent
User-agent: Amazonbot
Disallow: /en-us/support/announcement/information-regarding-cybernewsarticle
Disallow: /api/
Allow: /api/social-card
Disallow: /sign-in
Disallow: /sign-up
Disallow: /404
Disallow: /documents/*/legal/
Disallow: /documents/en-us/ParadigmB2B_Combine-26_vendor-Enterprise_VTEX.pdf
Disallow: /*/trust/cookies-information/
Disallow: /*/trust/platform/cookies-information/
# No blanket "Allow: /" here on purpose. Anything not disallowed is already
# allowed by default, and a parser that resolves conflicts by precedence rather
# than by specificity could read a blanket Allow as overriding the Disallow
# above it. The rule is only worth writing if it cannot be misread.

# See also: https://www.vtex.com/ai.txt