Agent Looker

We need to translate the given English text to Brazilian Portuguese, maintaining original format style. The user says: "Traduza seguinte texto para português Brasil, mantendo formato estilo original. Retorne EXCLUSIVAMENTE texto traduzido, sem qualquer comentário, aspas extras introduções:

[text]"

We need to output only the translated text, no extra commentary, no extra quotes. The original text is:

"Here's thinking process: 1. **Analyze User Input:** **Input Summary:** 'AI agents browse, click, take action. can't tell what trust. Agent Looker adds trust intelligence workflow keeps agents safe.' **Task:** Generate complete description 1000 characters. **Requirements:** Include emojis where appropriate make attractive. Focus main functionalities, benefits, use cases. Return ONLY plaintext (no HTML formatting). Language: Brazilian Portuguese. Return ONLY description text, filler phrases. **Constraints:** Max 1000 characters. 2. **Deconstruct Input Summary:** **Core idea:** agents perform webinteractions (browse, click, take action). Problem: Trust/verificationmissing ("can't tell what trust"). Solution: "Agent Looker" adds "trust intelligence" workflow. Benefit: Keeps agents safe. 3. **Identify Key Elements Description:** **Functionalities:** Web"

We need to translate this entire block to Portuguese, preserving formatting style (the original includes line breaks, bold markers, etc). However the instruction says "mantendo formato estilo original". So we should keep the same structure: line breaks, bold (**). But we need to translate the content, not the markup? The markup is part of the original style (maybe markdown). The instruction: "Traduza ... mantendo formato estilo original". So we translate the text, but keep the same formatting (including **). So we need to translate the English text inside the quotes and elsewhere, but keep the markdown bold markers.

Let's translate:

"Here's thinking process:" => "Aqui está o processo de pensamento:"

"1. **Analyze User Input:**" => "1. **Analisar Entrada do Usuário:**"

"**Input Summary:**" => "**Resumo da Entrada:**"

"'AI agents browse, click, take action. can't tell what trust. Agent Looker adds trust intelligence workflow keeps agents safe.'" => "'Agentes de IA navegam, clicam, realizam ações. não conseguem determinar confiança. Agent Looker adiciona fluxo de inteligência de confiança, mantendo os agentes seguros.'"

"**Task:** Generate complete description 1000 characters." => "**Tarefa:** Gerar descrição completa de 1000 caracteres."

"**Requirements:** Include emojis where appropriate make attractive. Focus main functionalities, benefits, use cases. Return ONLY plaintext (no HTML formatting). Language: Brazilian Portuguese. Return ONLY description text, filler phrases." => "**Requisitos:** Incluir emojis apropriados para torná‑lo atraente. Focar nas principais funcionalidades, benefícios e casos de uso. Retornar somente texto puro (sem formatação HTML). Idioma: Português do Brasil. Retornar somente o texto da descrição, sem frases de preenchimento."

"**Constraints:** Max 1000 characters." => "**Limitações:** Máximo de 1000 caracteres."

"2. **Deconstruct Input Summary:**" => "2. **Desconstruir o Resumo da Entrada:**"

"**Core idea:** agents perform webinteractions (browse, click, take action)." => "**Ideia central:** agentes realizam interações na web (navegar, clicar, agir)."

"Problem: Trust/verificationmissing ("can't tell what trust")." => "