fix(i18n): fix content mistranslations, garble, and brand-policy leaks (12 dicts)#219
Merged
Conversation
…s (12 dicts) PR 3 of the multi-agent re-review fix campaign (PR 1 #217 engine, PR 2 #218 data). Scope per owner decision: objective content DEFECTS + brand/product English- retention only. Stylistic word-choice (how to render "AI Fluency", subagent synonyms, generic rendering-consistency) is deferred to native review (#202). 108 value-only edits across the 12 source dictionaries, produced and then adversarially verified by an independent per-language pass (46 out-of-scope or incorrect proposals were rejected; every applied edit's prior value matched the file exactly; no keys added/removed, so key-parity holds): - Mistranslations (meaning was wrong): zh-CN Delegation 授权 "authorize" -> 委托 "delegate" and Diligence 审核 "review" -> 勤奋 "diligence" (the 4D competency names); zh-TW 審核 -> 勤勉; ja "steerable" 操舵可能 (nautical) -> 制御しやすい; ru "Headless mode" 自律 -> без интерфейса; it "Prompts" (verb-read) -> Prompt, trigger "grilletto" (gun trigger) -> attivazione, Sign In/Up swap. - Garbled strings (residue of the old unanchored protected-terms replaceAll): it "affidskill" -> affidabilità, "scalskill" -> scalabilità, "Aghook eventi" -> "Eventi degli hook". - Untranslated fragments: Bedrock/Vertex catalog connectives ("Claude with ..." -> con/avec/com), "Powered by", English lead-ins left mid-sentence (ko/ja). - Brand-policy: product names restored to English in prose — Skills / Agent Skills (the largest group: 技能/skill/Fähigkeiten/Habilidades/agent skills -> Skills/Agent Skills across agentSkills + claude101 + catalog), Claude Code (it "Codice Claude" -> Claude Code), Model Context Protocol (it). This pairs with PR 2, which removed the now-unsafe auto-restore of 技能/スキル/Skills — the curated values now carry the English brand directly. - German grammar: separable verb "Diese Lektion hervorhebt" -> "hebt ... hervor", genitive "Claude's" -> "Claudes", malformed compounds "KI-Fluencysplan" -> "KI-Fluency-Plan", "Lektion Rückblick" -> "Lektionsrückblick". 9 gates green · 553 unit · e2e 20/20 (one tight-timeout PDF-popup test flaked twice under local load; passed in isolation and on a determinism re-run — it is fixture-based and cannot be affected by dictionary content).
heznpc
added a commit
that referenced
this pull request
Jun 17, 2026
…policy (#221) An independent adversarial cross-validation (Codex, different model lineage) found two real defects in the recent i18n PRs; both reproduced locally. 1. Diligence regression (from #219). The AI Fluency "4D" competency Diligence is defined in-file as "informed responsibility for outputs — verify facts, check biases, ensure accuracy" (due diligence), but #219 changed zh-CN/zh-TW to 勤奋/勤勉 ("industrious/hardworking"), which mismatches that meaning. The pre-#219 审核/審核 ("review/verify") was actually closer. Revert all 14 occurrences (keeping the correct Delegation 委托/委派 fix). The optimal rendering (e.g. 审慎/尽责) is flagged for the #202 native review. 2. Agent Skills brand-policy was applied inconsistently in #219 — "Introduction to agent skills" still translated the product name in 8 locales. Restore the English product name "Agent Skills": es/fr/it/ja/pt-BR/vi/zh-TW + de hyphen (Agent-Skills → Agent Skills). ru is left as-is: its product term is the one genuinely contested cell (official RU docs translate Skills → "Навыки агента" while the dev community keeps English) — deferred to #202. Value-only edits (key-parity holds). 9 gates green · 553 unit · e2e 20/20. Two further Codex findings are design/minor and handled separately: the phonetic brand-transliteration restores (克洛德/クロード/Клод) corrupt real person names (Claude Monet/Debussy) — a real but low-frequency tradeoff on Academy pages, pending an owner decision on a name-particle guard; and an NFD/combining-mark edge where a precomposed protected form misses decomposed input (minor).
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
PR 3 of the multi-agent re-review fix campaign (#217 engine, #218 data).
Scope (owner decision): objective content defects + brand/product English-retention only. Stylistic word-choice (how to render AI Fluency, subagent synonyms, generic rendering-consistency) is deferred to native review (#202).
108 value-only edits across the 12 source dicts — produced then adversarially verified by an independent per-language pass (46 out-of-scope/incorrect proposals rejected; every applied edit's prior value matched the file exactly; no keys added/removed → key-parity holds).
Mistranslations (meaning was wrong)
Garbled (residue of the old unanchored protected-terms replaceAll)
it
affidskill→ affidabilità,scalskill→ scalabilità,Aghook eventi→Eventi degli hook.Untranslated fragments
Bedrock/Vertex connectives (
Claude with …→ con/avec/com),Powered by, English lead-ins left mid-sentence (ko/ja).Brand-policy (product names restored to English in prose)
Largest group — Skills / Agent Skills: 技能/skill/Fähigkeiten/Habilidades/agent skills → Skills/Agent Skills across agentSkills + claude101 + catalog. Also Claude Code (it "Codice Claude" → Claude Code) and Model Context Protocol (it). Pairs with PR 2, which removed the now-unsafe auto-restore of 技能/スキル/Skills — curated values now carry the English brand directly.
German grammar
Separable verb "Diese Lektion hervorhebt" → "hebt … hervor", genitive "Claude's" → "Claudes", malformed compounds "KI-Fluencysplan" → "KI-Fluency-Plan", "Lektion Rückblick" → "Lektionsrückblick".
Verify
9 gates green · 553 unit · e2e 20/20. (One tight-timeout PDF-popup test flaked twice under heavy local load; it passed in isolation and on a determinism re-run — it is fixture-based and cannot be affected by dictionary content.)
🤖 Generated with Claude Code