# Liftli — liftli.ai
#
# Policy: we WANT AI answer engines to crawl, cite AND learn from us (that is the
# point of /tools, /skills/ and llms-full.txt). Search, AI answers and AI training
# are all allowed (only Bytespider is blocked, below). Decided 2026-09-14:
# discoverability by agents over caution - the first paying customer came from an AI
# recommendation, and training is what puts a brand into a model's long-term memory. Content-Signal declares this without ever
# blocking the crawl; blocking the crawl would also block the citation.

User-agent: *
Content-Signal: search=yes,ai-input=yes,ai-train=yes
Allow: /

# Every AI crawler is allowed by the group above: answer engines (GPTBot, OAI-SearchBot,
# ChatGPT-User, ClaudeBot, Claude-User, Google-Extended, PerplexityBot) and training
# crawlers (CCBot, Amazonbot, Applebot-Extended, meta-externalagent). They deliberately
# have NO groups of their own: a crawler that matches a named group ignores this `*`
# group, and with it the Content-Signal line. The training crawlers were blocked until
# 2026-10-11, which contradicted ai-train=yes (external audit 2026-10-09).

# --- Blocked: ByteDance's crawler. Heavy crawl load, no citation or traffic back. ---
User-agent: Bytespider
Disallow: /

Sitemap: https://liftli.ai/sitemap.xml