On-brand content
Make every blog, deck, and post sound like your brand
One brand-voice skill, stacked with a content skill and a format skill, makes every blog, deck, email, and LinkedIn post sound like you. Learn skill stacking — the composition technique — and run it in Claude.
Why does AI content sound off-brand — and how do you fix it?
Ask an AI to “write a blog post” and you get competent, generic prose that sounds like everyone else — because the model has no idea what you sound like. Paste your style guide into every prompt and you’re copy-pasting the same instructions forever, and it still drifts.
The fix is to make your voice a reusable input instead of a repeated instruction. Skill stacking does exactly that: a voice skill produces a structured Brand Voice Card once, and every content skill reads it as an input. You stop describing your voice and start composing with it — which is why the output is consistent across every format, every time.
The framework behind it
The skill isn’t magic — it encodes a proven framework anyone can learn. Here’s the method itself, so you understand what it’s doing for you.
Skill stacking
the composition techniqueSkill stacking is the idea that skills compose. Instead of one giant do-everything prompt, you layer small, single-purpose skills so each does one thing well and hands its output to the next. The classic stack for content is three layers: a voice skill (how it sounds) × a content skill (blog, post, email) × a format skill (docx, pptx, pdf). Change any layer and the rest still works — swap the content skill and the voice stays; swap the voice and every format updates.
The reason it produces on-brand output is concrete, not hand-wavy: blog-generator’s own description says it requires a brand voice profile as input. So the voice skill isn’t decoration — it’s a literal input the content skill consumes. Generate a Brand Voice Card once, and every content skill you stack it with speaks in that voice. That’s how one voice skill makes every blog, deck, and post sound like the same brand.
The skills that get you there
Use any one on its own, or chain them — each hands its output to the next. Every one is free in the public library.
- 1
Brand Voice Generator
brand-voice-generator-v2The voice layer — produces a Brand Voice Card that every content skill reads as input, so everything sounds like you.
- 2
Blog Generator
blog-generatorThe content layer for long-form — it explicitly takes a brand voice profile as input, so blogs come out on-brand.
- 3
LinkedIn Post Generator
linkedin-post-generatorThe content layer for social — same voice, different format and length.
- 4
AEO Blog Generator
aeo-blog-generatorA content skill you can stack in for pages you want AI answer engines to cite — voice plus citability.
When you’d reach for it
Any time you publish across more than one format or channel and want it all to sound like one brand. The more surfaces you cover — blog, social, email, decks — the more a shared voice layer pays off.
- Marketing teams publishing across blog, LinkedIn, email, and decks
- Founders who are the brand voice and want to scale it beyond themselves
- Agencies producing content for clients who each need a consistent voice
- Anyone whose AI content reads as generic instead of “us”
What skill stacking does
It solves the “everything we publish sounds slightly different” problem. Capture your brand voice once, then stack it under any content skill — blog, LinkedIn post, email sequence, case study — and every output speaks in that voice. Add a format skill on top and it comes out as a doc, deck, or PDF, still on-brand.
The unit isn’t a single skill — it’s the stack. Voice × content × format, composed in one ask. It’s the same pattern our own marketing team runs, and it’s what turns a pile of skills into a repeatable content engine.
Try asking it
Real prompts, and what the skill hands back.
You ask
“Here’s our Brand Voice Card. Use it with the blog generator to write a 1,200-word post on prior-authorization automation.”
What comes back
An on-brand blog post — the content skill reads your voice profile as input, so it sounds like you, not like a generic AI.
You ask
“Same voice — now give me five LinkedIn posts and an email sequence from that post.”
What comes back
The same voice restacked under the social and email content skills — one brand voice, three formats.
Why stacking beats one big prompt
Voice as an input
Your Brand Voice Card is a literal input content skills consume — not an instruction you retype.
Composable layers
Voice × content × format. Swap any layer and the rest still works.
Consistent everywhere
One voice profile means the blog, the post, and the deck all sound like the same brand.
Single-purpose skills
Each skill does one thing well, so each is easy to improve — better than one brittle mega-prompt.
Dogfooded
It’s the exact pattern our own marketing team runs to stay on-brand at volume.
How the layers stack
The stack is voice × content × format: brand-voice-generator-v2 produces the voice profile; a content skill (blog-generator, linkedin-post-generator, aeo-blog-generator) consumes it; a format skill ships it. Compose them in one ask.
Scale this across a calendar and you get a content engine — see running a quarter of content through skill stacks.
Who it’s for
- Marketers who publish across many channels and want one consistent voice
- Founders scaling their own voice beyond what they can write themselves
- Content teams tired of AI output that reads as generic
Do it yourself — build your stack
Capture your voice once, then stack it under any content skill:
Point Claude at the right skills
agentman_skillsbrand-voice-generator-v2blog-generatorlinkedin-post-generatoraeo-blog-generatorAlready connected your library to Claude? Just tell it: “Use the brand-voice-generator-v2 skill (and the others below) from the agentman_skills MCP server.” Not connected yet? The Try in Claude / ChatGPT button above loads the first one from the public library — no setup. (connect your library.)
- 1
Load the voice skill
Click “Try in Claude / ChatGPT” above, or — if you’ve connected your library — tell Claude: “Use the
brand-voice-generator-v2skill from theagentman_skillsMCP server.” - 2
Generate your Brand Voice Card
Run
brand-voice-generator-v2to produce a structured voice profile — tone, vocabulary, rules. Save it; this is the input every content skill will read. - 3
Stack a content skill
Add
blog-generator(long-form),linkedin-post-generator(social), oraeo-blog-generator(citable pages) — and hand it your Brand Voice Card as input. - 4
Ask for the piece
Now ask for the blog, post, or page. Because the content skill consumes your voice profile, it comes out sounding like you — not like a template.
- 5
Add a format layer
Stack a Productivity skill (docx, pptx, pdf) on top to ship the piece as a doc, deck, or PDF — still on-brand.
- 6
Reuse the stack
Swap the content skill for the next format; the voice layer stays. One voice, every channel.
Build the brand voice once (see the [brand workflow](/showcase/skills/build-a-brand-from-scratch-in-claude)), and the whole content stack inherits it.
Try Brand Voice Generator right now
One click loads the skill into Claude or ChatGPT — no account, no setup.
FAQ
Brand Voice Generator: common questions
More in the showcase
Run a whole quarter of content from one calendar
Fan out each calendar entry into per-channel deliverables through skill stacks.
See it & try itBuild a brand from a blank page — the 7-step workflow you can run in Claude
Go from a blank page to a full brand book with seven skills, each feeding the next.
See it & try it