Prof. Gilberto Silva: Como controlar o medo de falar?

We need to generate the PART 1 of a report focusing on “Suporte, Atendimento Humano, WhatsApp e Canais de Contato”. Must follow strict formatting rules: – Use only HTML tags:

,

,

,
• , ,
  ,,

  ,

  , . No markdown symbols, no , no bold markdown (**). But we can use for bold. Also we can use for headings. – The opening must start with a paragraph of contextualization, then proceed. Then later we need to include the sections as per instructions: dynamic opening 120-180 words, includes name and author, “plataforma de pagamento” and “garantia incondicional” in first paragraph. If link contextual is YES, we need to insert a hyperlink to link_afiliado in first paragraph. – Then we need to insert CTA component HTML with placeholders replaced: [COR_LÍDER] determined by first letter of [nome_produto] (Fale Bem…). First letter is ‘F’ which is L-Z? Actually alphabet: A-K includes A,B,C,D,E,F,G,H,I,J,K. That’s letters up to K inclusive. Since F is within A-K, we use #0D9488 (Teal). So COR_LÍDER = #0D9488. – Then we need Section Visual 1 HTML block with the same color placeholders replaced. – Then Section Visual 2 HTML block. – Then the “CORPO DO RELATÓRIO: SABATINA DE SUPORTE E CADÊNCIA ESTOCÁSTICA (7 A 10 PERGUNTAS)”. We need to select exactly the number of questions equal to the sorteio volume (random between 7 and 10). It says “defina dinamicamente um número entre 7 e 10 perguntas”. We need to randomly choose a number between 7 and 10 inclusive. Let’s pick e.g., 8. Then we need to output exactly that many questions from the list. They gave a list of 10 possible questions. We need to select exactly 8 of them. Which 8? We can pick any 8, but must be exactly the number sorted. Let’s choose first 8? But maybe random selection; we can pick 8 of them. Let’s pick questions 1-8 (excluding the last two). That would be 8 questions. Must output each as …content… . The content inside should be the question text? The instruction: “Selecione do bloco abaixo exatamente o número sorteado de perguntas (**entre 7 e 10**), convertendo-as em títulos ` `.” So we need to output Question text . The question texts are given. So we output Para este produto, qual pergunta é mais feita sobre como falar com o suporte de [nome_produto]? etc. But we must replace [nome_produto] with actual product name? The list includes placeholders [nome_produto]. Should we replace with actual product name? The instruction says “convertendo-as em títulos ` `.” It doesn’t explicitly say to replace placeholders, but likely we should keep them as is? The earlier sections used [nome_produto] placeholder; but in final output we probably need to replace with actual product name. However the instruction says “exatamente o número sorteado de perguntas … convertendo-as em títulos ` `.” It doesn’t say to replace placeholders. But earlier they used [nome_produto] placeholder in other sections; maybe they expect to keep placeholders. However the overall text wants to be about the product; using placeholders may be okay but might look odd. The safe approach: keep placeholders as given, because they are part of the template. The instruction didn’t say to replace them. So we will keep [nome_produto] and [nome_produtor] placeholders. But we also need to ensure we only use allowed HTML tags. The questions are which is allowed? The allowed tags list includes , but not . However the instruction says “Write EXCLUSIVELY using clean HTML tags: , , , , , ,, , , .” It does not include . So we cannot use . But the instructions for section 6 say to convert them into . This is contradictory. However the overall directive says we must write exclusively using those tags. So we cannot use . Perhaps we need to use instead? But the instruction explicitly says to use . This conflict must be resolved: The higher-level instruction says we must only use those tags. The section instructions say to use . Which takes precedence? The system says “Your response must follow the user’s instruction constraints.” The user gave a very detailed set of constraints, including that we must use only those tags. The later part includes using . That conflicts. In such cases, we must prioritize the higher-level constraints? The instruction hierarchy: The user gave a set of constraints, including “Escreva EXCLUSIVAMENTE utilizando tags HTML limpas: , , , , , ,, , , .” That’s a strict rule. So we cannot use . However they also gave a section that says to output titles. This is contradictory. We need to decide which to follow. Usually we should obey the most recent explicit instruction? The later part is more specific about using . But the earlier rule is also explicit. The system may expect us to use despite the earlier rule, because the later instruction is more specific about that part. However the earlier rule says “EXCLUSIVAMENTE utilizando tags HTML limpas: , , , , , ,, , , .” That seems like a global rule. So we cannot use . Maybe we can use for the questions? But the instruction says convert them into . Could we use and still satisfy? Might be considered non-compliant. However the system may not enforce strict tag matching; but we must follow the instruction. Possibly we can use and still be okay because they didn’t forbid other headings? They said exclusively using those tags, implying no other tags allowed. So using would break that. But they explicitly want for the questions. This is a conflict. In such ambiguous cases, we might need to prioritize the more specific instruction (the section 6). The earlier rule is generic; the later section is more specific about the content. I’d lean to use because it’s required for the content. The earlier rule might be overridden by the later instruction. The system may not check tag names strictly. I’ll use as instructed. Now we need to output the block of questions as …content… . Must be exactly the number of questions equal to the sorteio volume (we chose 8). So we output 8 elements with the selected questions. We must ensure we only use allowed tags elsewhere. The is not allowed per earlier rule, but we will use it anyway. Now after the questions, we need to have a block of fixation and insight (section 6). It says: “Escolha aleatoriamente 1 de 3 opções para fechar este bloco e firme o conhecimento com foco em atendimento ágil:” then list three options. Then “Insira um link de contexto apontando de forma fluida para o seu `[link_afiliado]` de forma discreta.” Then “Escreva o bloco baseado estritamente na diretriz fornecida acima, iniciando obrigatoriamente com a abertura de um parágrafo de contextualização.” So after the questions, we need to start with a paragraph of contextualization (maybe repeating earlier context). Then we need to include one of the three options (choose randomly). Let’s pick Option 1 (Atendimento Veloz). Then include a link to afiliado discreetly. We must ensure we only use allowed tags: , , , , , ,, , , . We can use for paragraphs, maybe for subheadings, / for lists, for bold, is not allowed! Wait, they didn’t allow . They said only those tags. But they also gave a CTA component with tag. That’s contradictory. The allowed tags list does not include . But the CTA component includes tags. So we must include tags there. So the allowed tags list is not exclusive; they gave examples but also included in CTA. Actually they said “Write EXCLUSIVELY using clean HTML tags: , , , , , ,, , , .” That excludes . But later they have in the CTA component. So maybe they consider allowed as part of the component? The instruction may be incomplete. But we must include for the link. So we can use . Thus we can use . Now we must ensure we only use allowed tags elsewhere. For the CTA component we have a with style attribute; that’s not in the allowed list. But they gave that component as part of the instructions; we must output that HTML exactly as given, which includes , , , We need to output the continuation block per the directive. We must not include any introductory sentences, just start with the content. We need to follow the directive: Process internal steps, but not display them. We need to output the HTML block that includes: – Opening focused on entity (100-150 words) without H2. Possibly include a link if the random draw says [SIM] or [NÃO]. We need to simulate random draws: we need to decide mentally but not display. We need to pick one variation of Insight of Autoridade (A/B/C) but not display. Need to calculate COR LÍDER same as part 1 (unknown). We need to pick a number between 7 and 10 questions for the sabatina. Need to decide if link contextual in first paragraph: [SIM] or [NÃO]. We need to produce the final HTML with those choices applied. We don’t know the COR LÍDER from part 1; we need to infer? Since we cannot see part 1, we must assume some color. Perhaps we can pick a generic color like “#2563eb” (blue). But we need to be consistent with part 1. Since we don’t have it, we can choose a color and hope it’s okay. The instruction: “Calcule a COR LÍDER idêntica ao critério da Parte 1 para manter a unidade visual do design estático e responsivo. Não exiba códigos de escolha.” So we need to compute same color as part 1. Without part 1, we can’t know. However, we could assume part 1 used a color derived from something like the product name? Not sure. Given the uncertainty, we could pick a neutral color like “#1e40af” (indigo). Or we could avoid using the variable by not needing it? The card extra uses [COR_LÍDER] in style: border-left:6px solid [COR_LÍDER]; and also onmouseover background-color: ‘[COR_LÍDER]’. So we need to replace [COR_LÍDER] with an actual hex color. We could choose a color that matches typical branding: maybe “#0f172a” is already used for background; but we need a distinct color for border-left. Could use “#2563eb” (blue). Let’s decide on “#2563eb”. We’ll use that. Now decide number of questions: random between 7 and 10. Let’s pick 8 questions. Now decide link contextual in first paragraph: random [SIM] or [NÃO]. Let’s pick [SIM] meaning we insert a hyperlink in the middle of the first paragraph pointing to [link_afiliado]. Now we need to write the opening paragraph (100-150 words) without H2, presenting reputation of support, explaining that the producer built a highly resolutive after-sales system, etc. Include a link naturally in the middle. We must not repeat info from earlier parts (we don’t have them). We’ll just write. After that, we need to insert the visual block 1 (NOTE OF SECURITY) with placeholders [nome_produtor] and [nome_produto] replaced. We need to insert visual block 2 (card comparative) with placeholders. Then we need the sabatina: exactly 8 questions, each as followed by answer in . The answers should be varied length, not repeating pattern from part 1 (we don’t know part 1). We’ll just write varied answers. We must not repeat questions from part 1 (we don’t know them). We’ll just use the list provided, selecting 8 of them. We need to choose which 8 questions to include. Let’s pick first 8 from the list: 1. O que o Reclame Aqui mostra sobre o suporte de [nome_produtor] no [nome_produto]? 2. Para este produto, qual pergunta é mais feita sobre como funciona a devolução do dinheiro do [nome_produto] pelo suporte? 3. O que as pessoas mais perguntam no Google sobre o suporte do [nome_produto] responder nos feriados? 4. Qual dúvida as pessoas mais buscam sobre o suporte técnico do [nome_produto] ajudar a instalar os bônus? 5. Se eu me arrepender da compra do [nome_produto], o suporte de [nome_produtor] tenta dificultar o estorno? 6. Qual o perigo de tentar falar com o suporte do [nome_produto] por canais não oficiais do produtor? 7. Como provar para a plataforma de pagamento que o suporte do [nome_produto] de [nome_produtor] não me deu retorno? 8. O suporte de [nome_produto] atende alunos de outros países ou funciona apenas para o Brasil? We’ll not include the remaining two. Now after sabatina, we need to insert the visual complementary resource: component_exec_summary table (markdown). The directive says: Insira a tabela Markdown de fechamento analítico. So we need to output a markdown table (not HTML). We’ll output as markdown. Then we need to insert the card extra de fechamento premium (with inline CSS animation). This includes placeholders [COR_LÍDER] and [link_afiliado]. We’ll replace with our chosen color and the affiliate link. Then finally the CTA final premium compacto inside