edit-preserving-voicelisted

# Edit Preserving Voice Apply this skill as a language-quality layer over the current request. Let the content owner, tools, and evidence decide what belongs in the answer or artifact; shape only the generated language. Information selection and compression belong to `write-high-signal-update`; this Skill owns voice, register, locale, and fact-preserving expression. Complete the substantive request first, then apply the voice layer. Do not announce the skill, turn style guidance into a separate response, or add diagnosis unless the user asks for an audit, review, or explanation. ## Workflow 1. Derive substance from the request, evidence, and any composed domain skill. Treat voice as a delivery constraint, never a new source of truth. 2. Lock protected content before shaping language: facts, numbers, names, dates, citations, quotations, product/model names, code, commands, filenames, URLs, units, prices, benchmark conditions, legal/medical/financial caveats, author judgment, and uncertainty. 3. Identify the surface: chat reply, technical explanation, status update, documentation, release note, product copy, article, support message, translation/localization, or supplied-text transformation. 4. Infer only supported voice constraints from the request, conversation, or authorized sample: language, locale, register, directness, density, rhythm, politeness, code-switching, punctuation, and relationship to the reader. Ask only when a missing choice would materially change the r