# Cw Voice Check

> Diagnose whether writing matches the active writer, project, publication, and platform voice context, then offer a closer revision. Use when the user asks whether text sounds like them, feels generic, or has drifted from an established voice.

- Skill: `everyinc/cw-voice-check` (Agent Skill)
- Install (CLI): `npx skillmds@latest add everyinc/cw-voice-check`
- Raw SKILL.md: https://api.skillmd.com/api/skills/everyinc/cw-voice-check/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: everyinc (https://skillmd.com/u/everyinc)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/everyinc/cw-voice-check

---


# Voice Check

Evaluate the passage against the voice context that actually governs the assignment.

## Load Voice Context

Read `../../references/context-contract.md`, then load the relevant maintained voice and style sources it identifies.

Use this authority order for voice work:

1. Explicit instructions for the current passage.
2. The active writer and project `VOICE.md` files.
3. Project, publication, column, or platform `STYLE.md` and more specific guides.
4. Relevant curated examples from the active project's `examples/` folder.
5. Legacy `TASTE.md`, only when the project still maintains it.
6. Plugin defaults only for unresolved gaps.

Name the sources used in the assessment. Do not treat plugin defaults as the writer's identity.

## Diagnose

Check for:

- sentence architecture and cadence;
- diction and register;
- specificity, scenes, examples, and stakes;
- humor, vulnerability, priors, and distinctive asides;
- publication or platform fit;
- smoothing, generic symmetry, scaffolding, and other AI residue;
- phrases copied too closely from illustrative voice-guide examples.

Preserve intentional rule-breaking when it carries the writer's point of view.
Treat examples as evidence for voice rules, not a corpus to imitate.

## Output

Provide:

1. **Voice sources used** — the live files or supplied instructions.
2. **Aligned** — the strongest specific matches.
3. **Drift** — the few highest-impact mismatches, quoted precisely.
4. **Closer revision** — a usable rewrite that preserves meaning and authorship.
5. **Judgment call** — any tension between writer voice and publication/platform constraints.

Do not turn the result into a generic style checklist. The test is whether this piece sounds like this writer in this context.