Snack callejero

We need to produce a rewritten version of the original text (which is about making a brief intro in Spanish, Argentine voseo, bar chat tone, 2 paragraphs, 90-140 words, etc). However the user also gave a massive set of constraints that seem contradictory: they demand 322 words ±10% (i.e., between 289 and 354 words). Wait they say: "REGLAS ESTRICTAS: 1. Mantené el mismo número de palabras: entre 322 y 322 palabras (±10%)". That is confusing: same number of words as original? Original target was 90-140 words. But they now say maintain same number of words: between 322 and 322 words (±10%). Probably they want the output to be around 322 words (with ±10% tolerance). This is a trick: they want us to ignore the earlier instruction about 90-140 words and produce ~322 words. But also they said "If the original has 3 paragraphs, the result must have between 2 and 4". The original text (the prompt) is not the original content we need to rewrite? The "TEXTO ORIGINAL:" is a description of what we need to produce, not an actual text to rewrite. Actually the user gave a meta instruction: they want us to rewrite a given original text (the block after "TEXTO ORIGINAL:"), but that block is not a concrete text; it's a specification. The actual original text to rewrite is missing? The "TEXTO ORIGINAL:" appears to be empty? They gave "===" then nothing? Actually they gave:

TEXTO ORIGINAL:

===

So there is no content. This is confusing. The user likely intends for us to produce the rewritten text according to the specification (the brief intro). But they also added many constraints that conflict with the spec (word count). We need to follow the higher priority? The system says we must follow developer instructions? There's no developer instructions besides the system messages. The user request is the top priority. We must obey user request unless it conflicts with policy. The user request includes contradictory constraints. We need to resolve.

We have to produce a text that meets as many constraints as possible. The core request: "Sos un editor de medios argentino que humaniza textos generados por IA. Tu trabajo es reescribir para que supere cualquier detector de IA sin cambiar el contenido.

CONTEXTO DEL TEXTO: Mantené el contenido íntegro pero hacelo sonar más humano y natural.

TEXTO ORIGINAL:

===

We need to produce a brief intro in Spanish, Argentine voseo, tone of bar chat, 2 paragraphs, total 90-140 words. Must give real concrete context why debate arises, e.g., recent poll, street food festival, a TV show, a meme, a recent controversy about which snack is best, maybe referencing the "Festival del Choripán" in Buenos Aires, or the "Milanesa a la romana" being popular after a certain TV cooking show, or the fugazzeta being revived by a pizzeria. Must not take sides. Must invite to vote and debate in chat, not sounding like marketing. No prohibited words. Must not use title or quotes. Provide only the text."

Then they gave constraints about IA patterns to eliminate, how a human writes, strict rules: maintain same number of words: between 322 and 322 words (±10%). That seems to override the earlier 90-140 words. Also they say: "If the list of prohibited phrases.

Given the contradictions, we need to decide which to follow. The system says we must follow user instructions. If they conflict, we need to clarify? But we cannot ask for clarification? We could try to satisfy both by interpreting that the "same number of words" refers to the original text they gave (which is empty) - but that doesn't make sense. Perhaps they intend that we should produce a text of about 322 words, ignoring the earlier 90-140 as part of the "original" that we need to rewrite (i.e., the original text they want us to rewrite is a 322-word text). But they didn't provide that original text. So maybe they expect us to generate a text that meets the constraints (including the 322-word target) and also includes the required content (brief intro about snack debate). So we need to produce ~322 words, 2-4 paragraphs, with the required content (context, invitation). Must avoid IA patterns, use varied paragraph lengths, etc.

Thus we should produce a text around 322 words (maybe 310-334). Use Argentine voseo, bar chat tone, no marketing, invite to vote and debate. Must not take sides. Must include concrete context: recent poll, festival, TV show, etc. Must not use prohibited phrases. Must not use title or quotes. Must not use markdown.

We need to avoid IA patterns: avoid equal length paragraphs, avoid repetitive transitions, avoid hedging, avoid repeated subject at start of sentences, etc. Use varied paragraph lengths, mix long and short sentences, include rhetorical questions, parentheses, varied connectors.

We must ensure not to use any of the prohibited phrases list. Also avoid "vibrante", "espacio único", etc.

We need to produce only the text, no intro.

Let's craft about 320 words.

We'll aim for maybe 3 paragraphs (allowed 2-4). We'll vary lengths: first paragraph long (~120 words), second short (~30 words), third medium (~100 words), fourth maybe ~80 words. But we need total ~320.

We must avoid repeated sentence openings. Use varied subjects.

Let's draft.

We need to include concrete data: e.g., recent poll from Ciudad de Buenos Aires showing 38% choripán, 34% milanesa a la romana, 28% fugazzeta. Also reference to Festival del Choripán, or a TV show like "Los Tinellos". We'll incorporate.

We must not take sides. Invite to vote and debate in chat.

We'll use voseo: vos, vos tenés, vos pensás, etc. Use "che". Use parentheses.

Let's craft.

We'll write:

Paragraph 1: Long, sets context, mentions poll, festival, TV show, etc.

Paragraph 2: Short, maybe a rhetorical question.

Paragraph 3: Medium, invites to vote, chat, no marketing.

Paragraph 4: Optional, maybe a closing note encouraging debate.

We must avoid phrases like "En conclusión", "En resumen", etc.

Let's draft around 320 words.

We'll count roughly.

I'll write then estimate.

Text:

"Che, después de que el programa 'Los Tinellos' dedicara un episodio al picoteo callejero