ssheleg/super-ux / plugins/super-ux/skills/brand-voice
brand-voice
VerifiedUse when defining or checking how a product speaks: tone of voice, terminology, canonical facts, register. Triggers: "tone of voice" / "тон оф войс", "brand voice" / "голос бренда", "брендбук", "наша терминология". Not for writing the text (copywriting). Refusal phrase "no brand" / «без бренда».
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 7.8 KB |
| references/ai-tells.md | markdown | 20.6 KB |
| references/brand-contract.md | markdown | 25.4 KB |
| references/channel-playbooks.md | markdown | 20.0 KB |
| references/hooks.md | markdown | 26.4 KB |
| references/localization.md | markdown | 5.2 KB |
| references/marketing-copy.md | markdown | 8.1 KB |
| references/research-outliers.md | markdown | 5.4 KB |
| references/seo-aeo-safety.md | markdown | 9.9 KB |
| references/store-copy.md | markdown | 7.4 KB |
| references/surface-registers.md | markdown | 9.1 KB |
| references/system-map.md | markdown | 12.4 KB |
| references/video-script.md | markdown | 14.6 KB |
| references/voice-packs.md | markdown | 15.9 KB |
| templates/brand/README.md | markdown | 2.3 KB |
| templates/brand/channels.md | markdown | 4.2 KB |
| templates/brand/copy-brief.md | markdown | 187 B |
| templates/brand/facts.md | markdown | 1.3 KB |
| templates/brand/locale.md | markdown | 2.2 KB |
| templates/brand/strings.md | markdown | 1.7 KB |
| templates/brand/terminology.md | markdown | 1.2 KB |
| templates/brand/voice.md | markdown | 2.2 KB |
references/marketing-copy.md
View on GitHub# Marketing copy: pages, posts and long form
Craft for the marketing surfaces in [brand-contract.md](brand-contract.md).
Register per surface: [surface-registers.md](surface-registers.md). Platform
mechanics: [channel-playbooks.md](channel-playbooks.md). Anything a search
engine or answer engine will read also passes
[seo-aeo-safety.md](seo-aeo-safety.md).
## Contents
- [Before writing](#before-writing)
- [Page structure](#page-structure)
- [The seven sweeps](#the-seven-sweeps)
- [Word-level defaults](#word-level-defaults)
- [Long form: grounding](#long-form-grounding)
- [Comparison and competitor pages](#comparison-and-competitor-pages)
- [Blog and changelog](#blog-and-changelog)
Marketing copy is read once, by someone who did not ask for it, who is
deciding whether to keep reading. Every rule below follows from that.
## Before writing
For the current marketing writing task, record `docs/brand/copy-brief.md`
using the [brand contract](brand-contract.md#copy-briefmd). Reuse the current
brief only when its surface and action still match the task. Copywriting owns
this task record; it does not change voice, facts or approved targets.
The brief captures the action, reader and their words, proof, success metric,
constraints, and the reference to the voice's Admired/Refused pair. Write
`unknown` for unresolved inputs and report them at delivery. A missing metric
is not zero; a missing constraint is not permission; no brief author invents a
target, promise, fact or approval. Unknown proof never licenses a public claim.
Keep source links rather than copying the voice or its examples into the brief.
## Page structure
Above the fold: one headline, one subhead, one primary action.
Front-load the answer. The frequently quoted reading estimate comes from
[NN/g's 2008 analysis](https://www.nngroup.com/articles/how-little-do-users-read/)
of visit duration: 28% was a hypothetical upper bound if all time were spent
reading, with about 20% considered more realistic. It was not eye tracking or
a measurement of this product's audience. Use it as historical rationale for
scannable copy, never as a current conversion target or a universal quota.
| Section | Job |
|---|---|
| Headline | the single most important claim, checkable |
| Subhead | the specificity the headline could not carry |
| Primary CTA | verb plus outcome |
| Social proof | credibility, attributed |
| Problem | show you understand the situation |
| Solution | three to five outcomes, not features |
| How it works | reduce perceived complexity to three or four steps |
| Objections | comparison, FAQ, guarantee |
| Final CTA | recap, repeat, remove the risk |
**One idea per section.** A section that advances two arguments advances
neither; split it or cut one.
Headline shapes that survive contact: *outcome without the pain*
("Ship on Friday without the Friday deploy"), *category for an audience*
("The CI for teams that deploy hourly"), *never again* ("Never chase a flaky
test twice"), *the reader's own question*. Pick by what the traffic already
knows, not by preference.
### Page types
- **Homepage.** Several audiences without becoming generic. Lead with the
broadest true claim, then split the paths.
- **Landing page.** One message, one action, argument complete on the page.
The headline matches the ad or link that brought the reader; a mismatch is
the fastest bounce there is.
- **Pricing.** The anxiety is choosing wrong, not the number. Make the
recommended plan obvious, say what happens at the boundary between plans,
and state what happens at renewal.
- **Feature.** Feature to benefit to outcome, then a way to try it.
- **About.** Why the product exists, connected to what the reader gets. It
still carries an action.
- **Comparison.** See the honesty rule below.
## The seven sweeps
An editing pass, run in order. After each sweep, re-check the sweeps before
it: a fix in one dimension routinely breaks an earlier one.
1. **Clarity.** Can a stranger parse every sentence once? Jargon, ambiguous
pronouns, sentences carrying two ideas, points buried behind
qualifications.
2. **Voice and tone.** Consistent with `voice.md` and the surface's register?
Read it aloud; drift is audible before it is visible.
3. **So what.** For every claim, ask it literally. A feature with no
"which means…" bridge is a fact the reader has to convert themselves, and
they will not.
4. **Prove it.** Every claim carries evidence or gets softened. Numbers trace
to `facts.md`. Superlatives without an adjacent sourced fact are `B032`.
5. **Specificity.** "Save time" → "cut reporting from four hours to fifteen
minutes". Anything that cannot be made specific is usually filler, and
deleting it costs nothing.
6. **Emotion.** Does the before-state feel real? Pain named but not felt does
not move anyone. Emotion serves the argument; when it replaces the
argument, that is manipulation and it is out of bounds.
7. **Zero risk.** At the CTA: what is the reader afraid of, and does the page
answer it? Trial terms, cancellation, what happens to their data, what
happens next after the click.
## Word-level defaults
Cut: very, really, extremely, just, actually, basically, in order to.
Replace: utilize → use · leverage → use · facilitate → help · implement →
set up · robust → the specific number · seamless → name the step that
disappeared · innovative → say what is different.
Active voice. One idea per sentence. Vary sentence length deliberately:
uniform length is the most audible tell of machine drafting, and the fix is
rhythm, not vocabulary (see [ai-tells.md](ai-tells.md)).
No rhetorical dash and no full stop after a title. Both are `AT-06` and
`AT-07`, both are checked by `B062` and `B063`, and the replacement is
chosen from the meaning rather than by find-and-replace. Check the actual
heading after inline markup and responsive line breaks, plus any generator
that can restore old wording. Project display-copy rules may also cover hero
summaries and captions; paragraph punctuation stays normal. Keep semantic
heading roles in copy exports and put required `B063` checks in the delivery
gate with `--fail-on B063`, not an advisory-only command.
## Long form: grounding
A concept must be **grounded** before a later passage can lean on it. It is
grounded one of two ways: the reader walked in with it (a **prerequisite**),
or an earlier passage established it (**introduced**). A passage that reaches
for an ungrounded concept loses the reader, and the loss is silent.
The unit is the concept, not the word: a paragraph can lean on an idea the
reader lacks with no jargon in sight.
The lever is what you make a prerequisite versus what you ground inside the
piece. Demand too much up front and you shut out the readers you wanted;
ground too much inside and the opening drowns in definitions. Decide it
before the first paragraph, write it down, and when a section turns out to
need something ungrounded, the fix is a grounding passage before it or a
promotion to prerequisite, never proceeding and hoping.
Keep a running list of what is grounded. It is also the outline.
## Comparison and competitor pages
The one place copy is most tempted to lie, and where it is checked hardest.
- Compare against the competitor's **current, real** configuration, not a
straw one. Date the comparison and say when it was checked.
- Name at least one thing they do better. A comparison with no such line is
read as an advertisement and discounted entirely. The concession is what
makes the rest credible.
- Never quote a competitor's number that is not published by them.
- Where they are simply different rather than worse, say different.
## Blog and changelog
A blog post earns its place by carrying something the reader cannot get from
the six posts already ranking: original data, first-hand experience, or a
sharper frame. Restating the consensus in better prose is not a post.
Open with the point, not with context about the industry. Ground terms in
order. End where the argument ends.
A changelog entry says what changed **for the reader**. "Refactored the
scheduler" is a commit message; "Recurring jobs no longer drift by up to
90 seconds" is a changelog entry.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |