# VectoSolve - AI-Powered Image to Vector Converter
# https://vectosolve.com

# ============================================
# DEFAULT RULES — All standard crawlers
# ============================================
User-agent: *
Allow: /

# Allow Next.js static assets (Google needs CSS/JS for rendering)
# Disallow: /_next/static/css/
# Disallow: /_next/static/chunks/
# Disallow: /_next/static/media/
Disallow: /*?dpl=*

# Disallow private areas
Disallow: /lp/
Disallow: /api/
Disallow: /dashboard
Disallow: /dashboard/
Disallow: /auth
Disallow: /auth/
Disallow: /profile
Disallow: /_next/data/
Disallow: /admin
Disallow: /admin/
Disallow: /gestion
Disallow: /gestion/
# Espaces pilotes prives (une page par entreprise, acces par jeton)
Disallow: /pilot
Disallow: /pilot/

# Les images OG etaient bloquees ici pour economiser du budget de crawl. Le
# calcul etait faux: facebookexternalhit, LinkedInBot et Twitterbot respectent
# tous robots.txt, donc le blocage ne faisait pas economiser un crawl Google, il
# supprimait l'apercu de CHAQUE lien partage. Les pages /convert declarent
# /opengraph-image et le declaraient vers une image que personne n'avait le droit
# de lire, ce qui donne une vignette grise sur tout partage social.
# Une image de 1200x630 servie en cache n'est pas un poste de depense.

# Block PWA manifest
Disallow: /manifest.json

# Block duplicate query parameter URLs
Disallow: /blog?category=*
Disallow: /*?category=*

# Marketplace + shop + creators killed 2026-05-02 (pivot Niveau 2). All routes return
# 410 Gone with X-Robots-Tag: noindex. Deliberately NOT disallowed here: Googlebot must
# be able to crawl these URLs to SEE the 410 and drop them from the index. A Disallow
# would freeze them as "indexed, though blocked by robots.txt" (this is what kept them
# ranking until 2026-06-10).

# Inactive locales — middleware redirects /<inactive>/* to English with 308.
# Disallow ONLY the locales with zero measured GSC traffic. Locales bringing real
# clicks (ar/fa/id/uk/he/hr/vi/cs/pl/no — 50 clicks/28d in GSC) stay crawlable
# so their 308-redirect to English keeps serving users.
# Active locales (fr, es, de, pt, it, nl, ru, ja, ko, tr) stay crawlable.
# Decision-trigger: re-check GSC every quarter — promote any newly traffic-bearing
# locale to ACTIVE_LOCALES (real translations) or remove from this list.
Disallow: /zh/
Disallow: /zht/
Disallow: /hi/
Disallow: /th/
Disallow: /sv/
Disallow: /da/
Disallow: /fi/
Disallow: /ro/
Disallow: /hu/
Disallow: /el/
Disallow: /bg/
Disallow: /sk/
Disallow: /ms/
Disallow: /fil/
Disallow: /bn/
Disallow: /ta/
Disallow: /te/
Disallow: /mr/

# ============================================
# AI assistants — ALLOWED (decision 10/10/2026)
# GPTBot, ClaudeBot, Google-Extended, Meta-ExternalAgent and CCBot used to be
# blocked. Being known and cited by ChatGPT, Claude, Gemini and Meta AI is now
# a traffic channel worth more per visit than Google: they follow the default
# group above (private areas stay disallowed). ChatGPT-User, OAI-SearchBot,
# Claude-SearchBot, PerplexityBot, Applebot, YouBot and DuckAssistBot already did.
# ============================================

# Still blocked: scrapers and crawlers that feed no assistant our customers use.
User-agent: Bytespider
Disallow: /

User-agent: FacebookBot
Disallow: /

User-agent: cohere-ai
Disallow: /

User-agent: Diffbot
Disallow: /

# Sitemap location (split for reliable XML generation)
# 0 = static pages (tools, niche, landing pages)
# 1 = localized pages (homepages, tools, conversion)
# 2 = blog + translated blog articles
# 3 = SEO pages (conversion, comparison, alternative, tutorial, templates)
# Marketplace + shop product sitemaps removed 2026-05-02 (pivot Niveau 2)
Sitemap: https://vectosolve.com/sitemap/0.xml
Sitemap: https://vectosolve.com/sitemap/1.xml
Sitemap: https://vectosolve.com/sitemap/2.xml
Sitemap: https://vectosolve.com/sitemap/3.xml

# AI/LLM readable content — see https://llmstxt.org
# (LLMs.txt is non-standard; linked via /.well-known/ instead)
# https://vectosolve.com/llms.txt
# https://vectosolve.com/.well-known/llms.txt