Now encoding founding brands · BCP v0.1
AI agents are, or soon will be, writing your copy, buying your media, and describing your brand to customers. None of them have read your brand guidelines. brand.md fixes that.
Most companies will get that wrong. They'll treat brand as a cost center to automate past.
The tested brand knowledge that lives in senior marketers' heads. The judgment calls, the "we don't say it that way," the instinct for what's on-brand at the edge. None of it has a machine-readable home.
Without it, internal agents produce generic output, vendor platforms make decisions with partial information, and third-party consumer agents describe your brand however the model guesses.
The brands that win this cycle will be the ones whose strategy, voice, and judgment are encoded into the system itself. Authored once, read everywhere. That's what Encoded does.
In the world of agentic marketing, brand fidelity becomes an infrastructure problem, not a guidelines problem.
Same file. Different agents. Different moments. BCP is designed to be consumed inside the brand, across the vendor stack, and beyond the brand. No rewriting for each audience.
Copy, creative, media, compliance. Your internal agent stack consumes BCP to preserve senior-strategist judgment at machine scale. Every agent inherits the reflex on the first try.
Zefr, Canva, Adobe, Meta, The Trade Desk, DSPs, creative-gen tools. Their agents read BCP to make better calls on your behalf. Faster onboarding, sharper classification, fewer brand-safety misses.
ChatGPT, Claude, Perplexity, Gemini, agentic storefronts. BCP is the first time you get input into how third-party consumer agents describe your brand. You get input, not control.
Copy the prompt on the right. Paste it into Claude, ChatGPT, Cowork, or any AI you already use. It will interview you for about fifteen minutes, then hand you back a complete BCP tree ready to publish at your domain.
No signup. No account. No Encoded in the loop. Free forever. This is the whole point of an open standard.
Not technical? The prompt does the work. Show it to your AI of choice — Claude, ChatGPT, Cowork, Gemini, Perplexity — and it walks you through it.
A flat brand blob forces every agent to load everything to use anything. It's wasteful, and output quality degrades as attention dilutes. BCP is hierarchical: a lightweight brand.md at the root, daughters for voice, values, boundaries, claims, representation, audiences.
A media-buying agent pulls boundaries and claims. A copy agent pulls voice and audience. A creative-gen agent pulls visual and voice. Your brand shows up to each with exactly the context that task requires. Nothing more.
The file lives in a Git repo you own. We didn't build our own version control. GitHub handles versioning, diffs, history, access, and webhooks at enterprise grade, for free. We don't rebuild what already exists.
{domain}/.well-known/brand.mdPosition
The brand deck rotting on a shared drive is the antagonist of this decade.
Your agents can't read it. Your vendors won't. ChatGPT never will.
Encode the encoded. The craft your senior marketers already know. Put it into files every agent in the stack reads before it acts. Authored once. Consumed everywhere. Open, portable, yours.
A brand consultancy that can't encode its own brand into BCP has a credibility problem. So we ship ours in public, in Git, at encodedbrands.com/.well-known/brand.md. It is the reference template every brand forks.
Whatever patterns, structure, voice, and decisions we bake in will propagate across thousands of forks. We're writing the DNA of the standard, in public, with our own name on it.
---parent: /.well-known/brand.mdfile_type: voicelast_updated: 2026-04-17---# VoiceEncoded sounds like a senior brand strategist who hasstopped needing to prove anything and started needingto build something. Confident in the work, impatientwith noise, kind to the reader.## Preferprefer: - brand (as a noun, a verb, an object) - encode, encoded, encoding - agentic # not "AI-powered" - fidelity, living, ship## Avoidavoid: - AI-powered, AI-driven, AI-native # empty signifiers - revolutionize, game-changing, disrupt # hype - synergy, leverage, ecosystem # corporate - empower, unlock, supercharge # sludge## Do“The spec is open. The paid products are how wefund the work.”## Don't“We believe in the power of open standards combinedwith proprietary innovation to drive value at scale.”
Every path produces the same artifact: a versioned Git repo you own, published at your own domain, ready for every agent in your stack to read. Pricing from free to $15,000.
Forge, the self-serve wizard, is in development. Join the waitlist for early access.
Now encoding founding brands
We're encoding a small cohort of founding brands. Live platform integrations, co-announcement at launch, logo rights, hosting, versioning, and founding-customer status on paid tiers.