# robots.txt - lojaviaprofs.com (Loja Via Profs)
# Politica desta loja: busca e citacao nunca sao bloqueadas. Robos de busca,
# robos de IA de busca e citacao e robos de treino de modelos sao liberados
# por decisao documentada, nao por omissao. O que fica fechado para todos sao
# apenas areas tecnicas: painel, arquivos internos do WooCommerce (logs,
# temporarios e arquivos dos produtos vendidos) e links de adicionar ao carrinho.
# Guia para agentes de IA: https://lojaviaprofs.com/para-agentes-de-ia/
# Versao 2026-09-25 (arquivo fisico na raiz do site; substitui o robots.txt
# gerado pelo WordPress/WooCommerce)
#
# Atencao: um robo que tem grupo proprio (User-agent com nome) ignora o grupo
# "User-agent: *". Por isso todos os grupos abaixo repetem as mesmas protecoes.

# ------------------------------------------------------------------
# 1. Rastreadores em geral (busca tradicional incluida)
# ------------------------------------------------------------------
User-agent: *
Disallow: /wp-content/uploads/wc-logs/
Disallow: /wp-content/uploads/woocommerce_transient_files/
Disallow: /wp-content/uploads/woocommerce_uploads/
Disallow: /*?add-to-cart=
Disallow: /*?*add-to-cart=
Disallow: /wp-admin/
Allow: /wp-admin/admin-ajax.php

# ------------------------------------------------------------------
# 2. Guia para agentes de IA (onde esta loja fala com maquinas)
# ------------------------------------------------------------------
# pagina guia:   https://lojaviaprofs.com/para-agentes-de-ia/
# llms.txt:      https://lojaviaprofs.com/llms.txt
# llms-full.txt: https://lojaviaprofs.com/llms-full.txt
# brand facts:   https://lojaviaprofs.com/brand-facts.json
# sitemap de IA: https://lojaviaprofs.com/llms-sitemap.xml
# markdown de qualquer pagina: adicione /md/ ao final da URL
# MCP server:    https://lojaviaprofs.com/mcp (somente leitura, sem login)
# MCP card:      https://lojaviaprofs.com/.well-known/mcp/server-card.json
# catalogo:      https://lojaviaprofs.com/.well-known/ai-catalog.json
# ai.txt:        https://lojaviaprofs.com/.well-known/ai.txt
# agent skills:  https://lojaviaprofs.com/.well-known/agent-skills/index.json
# produtos JSON: https://lojaviaprofs.com/wp-json/wc/store/v1/products
# WebMCP:        ativo em todas as paginas (navigator.modelContext)

# ------------------------------------------------------------------
# 3. Robos de IA nomeados, liberados um a um
#    Busca e citacao (retrieval): OAI-SearchBot e ChatGPT-User (OpenAI),
#    Claude-SearchBot (Anthropic), PerplexityBot (Perplexity),
#    Bingbot (Microsoft Copilot), BraveBot (Brave Leo AI), ExaBot (Exa.ai).
#    Treino de modelos: GPTBot (OpenAI), ClaudeBot (Anthropic),
#    Google-Extended (token de opt-in do Google para IA).
#    Decisao: presenca no corpus e nas respostas vale mais que controle.
#    Os PDFs vendidos nao ficam expostos: sao entregues so ao comprador e a
#    pasta woocommerce_uploads segue fechada para todos.
# ------------------------------------------------------------------
User-agent: GPTBot
User-agent: OAI-SearchBot
User-agent: ChatGPT-User
User-agent: ClaudeBot
User-agent: Claude-SearchBot
User-agent: PerplexityBot
User-agent: Google-Extended
User-agent: Bingbot
User-agent: BraveBot
User-agent: ExaBot
Disallow: /wp-content/uploads/wc-logs/
Disallow: /wp-content/uploads/woocommerce_transient_files/
Disallow: /wp-content/uploads/woocommerce_uploads/
Disallow: /*?add-to-cart=
Disallow: /*?*add-to-cart=
Disallow: /wp-admin/
Allow: /wp-admin/admin-ajax.php
Allow: /

# ------------------------------------------------------------------
# 4. Demais robos de IA e de busca (assistentes, agentes e crawlers
#    de treino), todos com o mesmo acesso liberado
# ------------------------------------------------------------------
User-agent: anthropic-ai
User-agent: Claude-Web
User-agent: Claude-User
User-agent: CCBot
User-agent: Google-CloudVertexBot
User-agent: Perplexity-User
User-agent: Applebot-Extended
User-agent: Meta-ExternalAgent
User-agent: Meta-ExternalFetcher
User-agent: FacebookBot
User-agent: cohere-ai
User-agent: Bytespider
User-agent: DuckAssistBot
User-agent: MistralAI-User
User-agent: Amazonbot
User-agent: Timpibot
User-agent: GrokBot
User-agent: Grok-DeepSearch
User-agent: xAI-Bot
User-agent: YouBot
User-agent: DeepSeekBot
User-agent: AI2Bot
User-agent: PetalBot
User-agent: LinerBot
User-agent: QwenBot
User-agent: TongyiBot
User-agent: LinkupBot
User-agent: Slurp
User-agent: YandexBot
User-agent: Baiduspider
User-agent: Yeti
Disallow: /wp-content/uploads/wc-logs/
Disallow: /wp-content/uploads/woocommerce_transient_files/
Disallow: /wp-content/uploads/woocommerce_uploads/
Disallow: /*?add-to-cart=
Disallow: /*?*add-to-cart=
Disallow: /wp-admin/
Allow: /wp-admin/admin-ajax.php
Allow: /

# ------------------------------------------------------------------
# 5. Sinal de conteudo (Content Signals) e sitemap
#    ai-train=yes: conteudo publico liberado para treino de modelos.
#    search=yes: liberado para indexacao e busca.
#    ai-input=yes: liberado como insumo de respostas geradas por IA.
# ------------------------------------------------------------------
Content-Signal: ai-train=yes, search=yes, ai-input=yes

Sitemap: https://lojaviaprofs.com/sitemap_index.xml