Brand operations glossary

A long-lived credential scoped to specific brand kits and permissions, used by scripts and back-end integrations that can't run an OAuth flow.

API keys in Brand Kit OS are hashed at rest, scoped to least-privilege by default, and can be rotated or revoked without disrupting other integrations.

Related: MCP (Model Context Protocol), OAuth (Authorization Code with PKCE)

A character pattern (Hero, Sage, Outlaw, Caregiver, etc.) used to anchor brand personality and decision-making.

Archetypes give teams a shared shorthand for personality choices. They aren't a substitute for documented voice and governance — they're a lens that makes those rules easier to reason about and easier for AI models to apply consistently.

Related: Brand voice, Brand story

The gradual decay of brand consistency that happens when teams, agencies, and AI tools all interpret guidelines slightly differently over time.

Drift accelerates dramatically when AI tools enter the workflow because every prompt is a chance to interpret the brand. Structured brand kits plus governance reviews are the standard mitigation.

Related: Brand governance, Diligence framework

The rules, approvals, and review processes that keep brand work on-strategy as teams, agencies, and AI tools scale.

Governance answers: who can change brand-defining fields, what gets reviewed before it ships, and how AI-generated content is checked for compliance. In Brand Kit OS, governance is structured data — not a Slack channel — so it can drive automated diligence checks.

Related: Compliance standard, Diligence framework, Role (admin, author, viewer)

A single source of truth for everything a brand stands for — voice, audience, visual identity, governance — captured as structured data instead of a static PDF.

In Brand Kit OS, a brand kit groups every brand attribute under one workspace so humans and AI agents both pull from the same fields. It replaces the traditional brand-guidelines document with machine-readable data that can be exported to ChatGPT, Claude, MCP-enabled agents, and Markdown briefs without rekeying.

Related: Brand voice, Visual identity, Brand governance, Export

The narrative that explains why a brand exists, who it serves, and what change it wants to make in the world.

Brand story is the source material that everything else (positioning, voice, campaigns) is derived from. Capturing it explicitly in a brand kit means AI tools and new team members can ground their work in the same narrative instead of inferring it.

Related: Messaging framework, Brand archetype

The consistent way a brand sounds — tone, vocabulary, sentence rhythm, and the things it would never say.

Brand voice is captured as a structured set of attributes (tone descriptors, do/don't lists, example phrasings, negative directory) rather than free-form prose. This lets AI tools enforce voice deterministically instead of guessing from a paragraph of guidelines.

Related: Tone, Messaging framework, Negative directory

A packaged bundle of instructions, agents, and reference files that extends Claude with brand-specific behavior.

Brand Kit OS exports Claude skill bundles that load brand voice, governance rules, and review agents into Claude with a single drag-and-drop. The skill keeps using the latest brand data on every run.

Related: Export, MCP (Model Context Protocol)

A specific brand or regulatory rule that content must satisfy — for example, claims language, accessibility, or industry-specific disclosures.

Compliance standards are individual checkable rules attached to a brand kit. Exports and AI workflows can reference them by ID so a single source change updates every downstream guardrail.

Related: Brand governance, Diligence framework

A repeatable checklist Brand Kit OS uses to evaluate brand work across Creation, Transparency, and Deployment dimensions.

The framework turns governance from a vibe into a scoring system. It surfaces gaps before content ships and produces evidence trails for clients, regulators, or internal review.

Related: Brand governance, Compliance standard

Packaging a brand kit into the exact format a specific AI tool, agent, or workflow expects — for example a ChatGPT GPT, a Claude skill bundle, or a Markdown brief.

Exports are the bridge between structured brand data and the AI tools where work actually happens. Brand Kit OS supports one-click exports so updating the source brand kit propagates to every downstream surface without manual re-entry.

Related: Brand kit, MCP (Model Context Protocol), Claude skill

How a brand shows up across channels — the applied combination of voice, visual identity, and tone for a specific surface.

Expression bridges strategy and execution. A brand might have one identity but distinct expressions for product UI, marketing campaigns, and sales collateral, each governed by the same underlying brand kit rules.

Related: Visual identity, Brand voice

A document uploaded to a brand kit (PDF, Markdown, transcript, deck) that AI tools can reference as ground truth.

Knowledge files extend a brand kit with longer-form context that wouldn't fit in structured fields — research, interview notes, internal playbooks. They're exported into AI tools alongside the structured brand data.

Related: Brand kit, Export

An open protocol that lets AI agents (Claude, Cursor, and a growing list of clients) call external tools and read external context in a standard way.

Brand Kit OS ships a hosted MCP server so any compatible client can query a brand kit live — without re-uploading exports each time something changes. Authentication uses OAuth 2.0 with PKCE or scoped API keys.

Related: OAuth (Authorization Code with PKCE), API key, Export

A structured map of what a brand says, to whom, and why — usually positioning statement, value pillars, proof points, and audience-specific messaging.

Messaging frameworks turn strategy into reusable building blocks. In Brand Kit OS each pillar is a field that can be referenced by exports, AI prompts, and persona briefs so a launch page and a sales email both inherit the same positioning.

Related: Brand story, Target audience, Value pillar

An explicit list of words, phrases, claims, or topics the brand never uses.

Negative directories prevent off-brand drift in AI-generated content by giving the model an exclusion list as concrete as the inclusion list. They're often the single highest-impact addition to an AI brand kit.

Related: Brand voice, Brand governance

The authentication flow Brand Kit OS uses for remote MCP connectors — the user approves a client once and the client gets a scoped token to call the API.

PKCE (Proof Key for Code Exchange) protects the flow against intercepted authorization codes. It's the recommended pattern for public clients like desktop AI tools that can't safely hold a long-lived secret.

Related: MCP (Model Context Protocol), API key

A concrete, named representative of a target audience used to guide content, design, and product decisions.

A good persona captures jobs-to-be-done, objections, and information sources — not just demographics. Personas live alongside the audience definition in a brand kit and are exported into AI tools as first-class context.

Related: Target audience, Messaging framework

An attack where untrusted text inside an AI tool's input overrides the original instructions and changes the model's behavior.

Brand Kit OS mitigates this by treating brand context as system-level data and validating exports server-side so adversarial blog posts or knowledge files can't escalate into instruction changes.

Related: Brand governance

A Postgres feature that enforces who can read or write each row in a table — the foundation of multi-tenant data isolation in Brand Kit OS.

Every user-owned table in Brand Kit OS has RLS policies that scope reads and writes to the brand kits a user actually belongs to. Service-role workloads (edge functions, admin tools) bypass RLS deliberately and apply their own access checks.

Related: Role (admin, author, viewer), Brand governance

Permission level for a user inside a brand kit — admins manage settings and members, authors edit brand fields, viewers read only.

Roles in Brand Kit OS live on the brand kit membership, not the user record, so the same person can be admin of one brand and viewer of another. Platform staff have a separate flag that grants cross-account access for support.

Related: Brand governance

The specific group a brand exists to serve — defined by attributes like role, goals, pain points, and buying signals.

Target audience is the broad cohort; personas are concrete representatives of that cohort. Both belong in a brand kit because AI tools produce dramatically better output when they know who they're writing for.

Related: Persona, Messaging framework

How brand voice shifts by context — the difference between an apology email and a launch announcement.

Tone modulates voice for a specific moment, channel, or audience. A brand can have one voice and many tones; capturing tone separately keeps writers and AI tools from flattening every piece of content into the same register.

Related: Brand voice, Messaging framework

A short, defensible statement of value that a brand returns to repeatedly across messaging and product positioning.

Value pillars sit inside the messaging framework. They're the most-quoted source for AI tools when drafting landing pages, sales emails, or campaign concepts, so they reward careful wording.

Related: Messaging framework, Brand story

The visual system that makes a brand recognizable — logo, color palette, typography, imagery rules, and layout principles.

In an AI-native brand kit, visual identity is captured as structured tokens (hex codes, font families, usage rules) rather than image references. That lets design tools, code, and AI image generators all pull from the same authoritative values.

Related: Brand kit, Expression