Help us improve
Share bugs, ideas, or general feedback.
From skills-for-humanity
Extracts a voice fingerprint from strong passages to audit and repair voice departures in multi-author documents or when brand voice has drifted.
npx claudepluginhub human-avatar/skills-for-humanityHow this skill is triggered — by the user, by Claude, or both
Slash command
/skills-for-humanity:s4h-writing-voice-consistencyThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Voice is the sum of choices: sentence length preference, vocabulary range, degree of formality, stance toward the reader, use of metaphor and analogy, tolerance for digression, the character that comes through even in functional writing. It is recognisable even when you can't name what you're recognising. When it's absent or fractured, the reader feels it as a nagging sense that the writing has...
Extracts, codifies, and replicates brand voice from sample copy. Auto-activates on copywriting tasks to protect client tone from generic AI prose.
Diagnoses and repairs tone drift across dimensions like formality, warmth, urgency, stance, and sentence rhythm. Useful for multi-author docs, long pieces, or source-assembled content with inconsistent voice.
Detects unexpected tone or register shifts in writing, offers rewrite suggestions to restore consistent voice. Useful for multi-author drafts or branded content.
Share bugs, ideas, or general feedback.
Voice is the sum of choices: sentence length preference, vocabulary range, degree of formality, stance toward the reader, use of metaphor and analogy, tolerance for digression, the character that comes through even in functional writing. It is recognisable even when you can't name what you're recognising. When it's absent or fractured, the reader feels it as a nagging sense that the writing has no centre — that no one is actually speaking to them.
The critical insight: voice can be extracted and characterised from strong existing passages, and that characterisation becomes an auditable standard. This means voice consistency is not a mysterious quality that can only be felt — it is a set of identifiable, describable choices that can be held constant across contributors, sections, and time.
For brand and organisational writing, this extracted voice fingerprint becomes the foundation of a practical style brief — not a list of rules about comma usage, but a description of who is speaking and how they sound, with examples that any contributor can use as a reference.
Step 1: Extract Defining Characteristics From the strongest existing passages — the sections that feel most alive, most like the brand or the writer's best work — identify 5–7 defining characteristics. These should be specific enough to be testable. Not "clear and friendly" but:
Framing check: Confirm the specific voice and document before continuing. State what you've identified — the actual text being analyzed, whether this is a single-author or multi-author document, and the primary goal (audit, repair, or guide creation) — in one sentence, then use AskUserQuestion:
Before narrowing: Show the complete generated set of characteristics to the user first. Use AskUserQuestion:
Step 2: Create a Voice Fingerprint Write a representative passage (or select the strongest existing one) that exemplifies all 5–7 characteristics simultaneously. This is the voice fingerprint — the reference point that any subsequent section can be held against. It should feel unmistakably like the voice.
Step 3: Scan for Departures Read through the full document and flag sections that depart from the fingerprint. For each departure:
Step 4: Build a Voice Guide (for brand/multi-author use) For documents that will have multiple contributors or ongoing updates, translate the fingerprint into a brief, practical guide:
Before proceeding, use the AskUserQuestion tool. State your interpretation of the situation in 1–2 sentences — what is being analyzed and what the core question is — then ask:
Proceed based on their selection. If the user reframes, incorporate the correction before running any analysis.
Fingerprint (5–7 Characteristics):
Representative Sample Passage: [The passage that exemplifies all characteristics — quoted from existing text or composed if no strong existing passage exists]
Departures:
Voice Guide (for brand/multi-author use):
/s4h-writing-tone-alignment — tone is one dimension of voice. Voice consistency addresses the full identity of the speaker; tone alignment addresses one specific dimension (register)./s4h-writing-audience-calibration — voice must be calibrated to the audience; a perfectly consistent voice that is calibrated to the wrong reader is still failing./s4h-writing-prose-elevation — elevation must stay in voice; an elevated passage that sounds like a different writer has traded one problem for another.After delivering this output, use AskUserQuestion to offer the next move:
/s4h-writing-line-editing — Edit for consistent voice throughout/s4h-writing-tone-alignment — Align tone with the voice/s4h-writing-prose-elevation — Elevate the consistent voice