git:20260918.40e721d to git:20261013.a69e0bb

19 added, 7 removed. Audit A to A.

---
name: content-locale-humanize-pt
title: "Portuguese Native-Fluency Calibration"
description: "Per-language calibration anchors for detecting AI-slop and translationese in Portuguese (pt) target text; auto-loads alongside content-locale-humanize when the target language is Portuguese."
license: Apache-2.0
- compatibility: "Extends content-locale-humanize's AI-Slop & Translationese dimension. Load both together — this file has no rubric of its own."
+ compatibility: "Calibration data only; no tools required. Network access helps for usage checks."
domains: content
rules:
- content(portuguese)
- match(\bpt-(BR|PT)\b)
- match(\bPortuguese\b)
- match(português)
- semantic(check if this Portuguese translation sounds native)
- semantic(eliminate translationese from this Portuguese text)
---
## Overview
- Sourced calibration anchors for Portuguese, feeding `content-locale-humanize`'s AI-Slop & Translationese dimension. This is a calibration aid, not the checklist — reason natively beyond it (see the core skill's "why structure, not word lists" section).
+ Calibration anchors and current norms for Portuguese (pt-BR, pt-PT), feeding content-locale-humanize's dimensions. A calibration aid, not the checklist; reason natively beyond it.
- ## Instructions
+ ## Rules
- Sourcing confidence: moderate — listicle-level convergence (multiple independent sites), not corpus research.
+ ### AI tells
- Overused words: crucial, jornada (calque of English AI-cliché "journey" as a metaphor, not literal travel).
+ Sourcing confidence: moderate; listicle-level convergence across independent sites, not corpus research.
- Watch for the same connector-pileup and stock-opener patterns documented for Spanish and French — Portuguese AI output leans on the same "além disso"-class transitions. Distinguish pt-BR from pt-PT register and vocabulary; don't apply European Portuguese conventions to a Brazilian-audience piece or vice versa.
+ - Overused words: crucial, jornada (calque of the English AI metaphor "journey").
+ - The same connector and opener patterns as Spanish and French: "além disso", "vale ressaltar", "é importante destacar", "em suma".
+ ### Translationese from English
+
+ - pt-BR gerundismo ("vou estar enviando" for "will be sending"; use "vou enviar").
+ - False friends: "aplicar para uma vaga" (candidatar-se), "realizar" for perceber, "eventualmente" meaning finally, "suportar" for ser compatível com.
+ - Possessive overuse ("seu/sua" everywhere), "foi feito por" passives, Title Case headings.
+
+ ### Current norms
+
+ - Never mix variants. pt-BR: você, "estou fazendo", celular, tela, usuário, arquivo, “…”. pt-PT: tu or você by register, "estou a fazer", telemóvel, ecrã, utilizador, ficheiro, «…».
+ - Spelling follows the 1990 Orthographic Agreement, mandatory in Brazil and Portugal; residual differences remain (pt-BR "fato", pt-PT "facto").
+
## References
- humanizartextos.com, gowinston.ai.
+ Reviewed 2026-10-12. AI tells: humanizartextos.com, gowinston.ai. Norms: Vocabulário Ortográfico da Língua Portuguesa (ABL; Instituto de Lexicologia e Lexicografia da Língua Portuguesa).