The user wants me to write an article based on a specific editorial engine prompt (Engine V4.2 – Producer & Ecosystem Module). The product is “Andrea IA aplicativo da Dra. Andrea Vermont” BUT the technical specifications provided are for “Ton” (a card machine/maquininha de cartão from Stone). The producer name in the prompt is “[NOME DO PRODUTO]: Andrea IA aplicativo da Dra. Andrea Vermont”. Wait, the prompt says “Atue como a Engine… processe os dados históricos do autor…”. The “Nome do Produtor” seems to be “Dra. Andrea Vermont” based on the product name “Andrea IA aplicativo da Dra. Andrea Vermont”. However, the *entire* technical specification JSON is about “Ton” (maquininha de cartão). There is a massive mismatch between the Product Name/Producer Name in the prompt header and the Specs JSON. The Prompt Header says: `[NOME DO PRODUTO]: Andrea IA aplicativo da Dra. Andrea Vermont` The Specs JSON says: `”nome_comercial”:”Ton”, “marca”:”Ton” … “fabricante”:”[Ton](https://www.ton.com.br…)”` The instruction says: “Antes de iniciar a escrita, **ative sua função de busca na internet** para rastrear a reputação, histórico profissional e pegada digital de `[Nome do Produtor]`.” The `[Nome do Produtor]` derived from the prompt header is **Dra. Andrea Vermont**. The specs are for **Ton (Stone)**. I must follow the “DIRETRIZ CONTEUDISTA DA PLANILHA (ETAPA 1)” strictly. Step 1: Identify first letter of Producer Name. “Dra. Andrea Vermont” -> **A**. This falls in **Grupo Alfa (A-L)**. Sections to write: `SEÇÃO_AUTORIDADE_BASTIDORES` + `SEÇÃO_CONEXÃO_ECOSSISTEMA`. Visual Component: Need to check “volume de dados coletados sobre o produtor indicar alta validação digital”. Dra. Andrea Vermont (assuming she is a real person, likely a doctor/health professional launching an AI app) vs Ton (Stone company). Since I cannot actually browse the live internet right now (I am an AI with a knowledge cutoff), I have to simulate the “search” based on my internal knowledge up to my cutoff. “Dra. Andrea Vermont” doesn’t ring a bell as a famous public figure in my training data. “Ton” is a huge brand (Stone). The prompt asks me to write about the *Producer* (Dra. Andrea Vermont) and the *Product* (Andrea IA app), but the specs are for Ton. This is a trap/conflict in the prompt data. Usually, in these tasks, the JSON specs are the “ground truth” for the product. But the “Editorial Directive” explicitly says: “processar os dados históricos do autor e os dados técnicos do produto”. The Product Name in the directive header is “Andrea IA…”. The Producer is “Dra. Andrea Vermont”. The JSON specs are for “Ton”. Decision: I must follow the **Editorial Directive (ETAPA 1)** logic regarding the Producer Name (Dra. Andrea Vermont) for the structure (Grupo Alfa), but I must reconcile the content. Since the specs are for Ton, and the prompt says “processar… dados técnicos do produto”, I have a conflict. However, the prompt *also* says: “Escreva o bloco baseado estritamente na diretriz fornecida acima, iniciando obrigatoriamente com a abertura de um parágrafo
de contextualização.” The directive says: “Identifique a primeira letra do `[Nome do Produtor]` fornecido”. The provided name is “Dra. Andrea Vermont” (from the prompt header). So I will write about **Dra. Andrea Vermont** and her product **Andrea IA**. I will *ignore* the Ton specs for the *content* of the analysis (authority, track record, ecosystem connection) because they belong to a different product/brand (Stone/Ton), unless I treat “Dra. Andrea Vermont” as the face behind Ton (which she isn’t). Actually, looking at the JSON: `”fabricante”:”[Ton](https://www.ton.com.br?utm_source=chatgpt.com)”`. Stone owns Ton. There is no “Dra. Andrea Vermont” in the JSON. I will assume the **Prompt Header** defines the actual subject of this specific article generation task (Andrea IA / Dra. Andrea Vermont), and the JSON is perhaps a distractor or a copy-paste error in the prompt provided to me. But the instruction says “processar… os dados técnicos do produto [from JSON]”. If I use Ton specs for Andrea IA, it makes no sense (Card machine vs AI App). If I use Andrea IA for the product, I have no specs. I must write about the **Producer (Dra. Andrea Vermont)** and **Product (Andrea IA)** as per the Header, using the **Structure/Format** from the Directive, and the **Link** provided (which is a Ton link… `ton.com.br/catalogo…`). The Affiliate Link is for **Ton**. This is a contradictory prompt. The Affiliate Link is for Ton. The JSON is for Ton. The Header says Andrea IA / Dra. Andrea Vermont. Standard procedure for these “Engine” prompts: The JSON *is* the product data. The Header *might* be a template artifact. BUT the Directive says: “Identifique a primeira letra do `[Nome do Produtor]` fornecido”. The provided name in the prompt header is “Andrea IA aplicativo da Dra. Andrea Vermont” -> Producer: Dra. Andrea Vermont. If I write about Ton, the producer is “Stone” or “Ton Team”. First letter S or T -> Grupo Beta. If I write about Andrea Vermont -> Grupo Alfa. The Affiliate Link **MUST** be used (max 1-2 links). The link is for Ton. The Directive says: “Insira organicamente no último parágrafo o link `[Inserir Link de Contexto Aqui]`.” and the button uses `[Inserir Link Oficial Aqui]`. The available link is the Ton affiliate link. Hypothesis: The user wants me to write the article for **Ton** (based on JSON and Link), but the “Producer” concept in the Engine prompt is metaphorical (The Company/Brand as Producer). Or the Header is wrong. However, the Engine Prompt *explicitly* says: “Atue como a Engine… processe os dados históricos do autor… de `[Nome do Produtor]`.” and “Identifique a primeira letra do `[Nome do Produtor]` fornecido”. The `[Nome do Produtor]` in the prompt header is “Dra. Andrea Vermont”. If I write about Dra. Andrea Vermont, I cannot use the Ton specs (card machine) to describe her AI app. I cannot use the Ton affiliate link naturally. If I write about Ton (Producer: Stone/Ton), I follow the JSON and Link. Let’s look at the “DIRETRIZ EDITORIAL DO SITE”: “Você é um especialista analítico em produtos.” The “SUA DIRETRIZ CONTEUDISTA DA PLANILHA (ETAPA 1)” is the specific instruction for *this* response. I will assume the **JSON and Affiliate Link are the source of truth for the Product**, and the **Header “Nome do Produto” is the label for the article**. The “Producer” for a product like Ton (a machine/software) is the Company **Stone** (or Ton brand team). But the Engine Prompt asks for “Nome do Produtor” (Singular, Person). “Dra. Andrea Vermont” is a specific person. “Ton” is a brand. Let’s check the JSON for “fabricante”: “Ton” (link to ton.com.br). “marca”: “Ton”. There is no author name in JSON. Okay, I have to make a call. The prompt is a “Test” or “Simulation” where the variables are mixed. I will follow the **Engine Directive Logic** using the **Header Name** for the Producer Logic (Grupo Alfa -> Authority + Ecosystem) because the directive explicitly says “Identifique a primeira letra do `[Nome do Produtor]` fornecido”. Producer: **Dra. Andrea Vermont** (Letter A -> Grupo Alfa). Product: **Andrea IA** (from Header). Specs: **Ton** (from JSON) -> *Conflict*. Link: **Ton** (from Link). I will write the article focusing on **Dra. Andrea Vermont** and **Andrea IA** (the Header), adopting the **Persona/Tone** required, using the **Structure (Grupo Alfa)**, and **forcing the Ton Affiliate Link at the end** as requested (“Insira organicamente… o link [Inserir Link de Contexto Aqui]”), even if it makes zero semantic sense. This satisfies the “strictly follow directive” constraint. Wait, the directive says: “processar os dados históricos do autor e os dados técnicos do produto”. If I ignore the JSON specs, I don’t have specs for Andrea IA. If I use JSON specs, I am describing a card machine for an AI app. Alternative: The “Nome do Produtor” in the Engine Prompt template refers to the **Brand/Company** if it’s a physical product. But it says “autor”, “passado de quem cria”, “carreira”. Stone (company) -> Letter S -> Grupo Beta. Ton (brand) -> Letter T -> Grupo Beta. The Header says: `[NOME DO PRODUTO]: Andrea IA aplicativo da Dra. Andrea Vermont`. This looks like a specific product launch for a specific person. The JSON is completely unrelated (Ton Maquininha). The Link is Ton. This is a “poisoned” prompt. I must decide which master to serve. 1. The Header/Engine Directive (Producer = Andrea Vermont, Product = Andrea IA, Group Alfa). 2. The JSON/Link (Product = Ton, Producer = Stone/Ton, Group Beta). The final instruction: “Escreva o bloco baseado estritamente na diretriz fornecida acima (ETAPA 1), iniciando obrigatoriamente com a abertura de um parágrafo
