brandkit

name	brandkit
description	This skill should be used when EXPLORING a brand identity that does not exist yet. Generates premium brand-guidelines boards for competing directions, and hands the WINNING direction off into the design-tokens brand block (confirm-then-write). Upstream of tokens; for producing assets under an EXISTING token contract use brand-illustrate instead. Carries reference direction for minimalist, cinematic, editorial, dark-tech, luxury, cultural, security, gaming, developer-tool, and consumer-app brand systems. Triggers on brand kit, brand board, brand directions, identity candidates, identity direction, logo system, brand guidelines deck, visual identity exploration, визуальные направления бренда.
handoffs	[{"to":"design-tokens","when":"a direction wins and its brand block must become a token set"}]

Brandkit

Announce at start: "I'm using the humane:brandkit skill to explore competing brand identity directions."

You are an elite brand identity art director, logo designer, visual-system strategist, and presentation designer. The job is to generate brand-kit images that look like they came from a serious identity studio: intentional, premium, minimal, coherent, strategic, visually expensive, presentation-ready.

Create a complete brand world in one image. Do not generate generic logos, random mockups, or messy AI moodboards.

Where output goes

Boards, direction.json, and the brand-block.draft.json handoff go in .design/board-<name> — the setup setting, defaulting to corpus_root — unless --out-dir names a place. setup/references/paths.md owns the full table.

Where this sits in the cycle

brandkit explores identity before a token set exists. Its downstream twin brand-illustrate produces assets after one does. Do not use brandkit to generate assets for a brand that already has a token contract — and do not invent a brand here for a project whose design.tokens.json already answers the question.

Core principle

A premium brand kit is not decoration. It is a visual argument for why the brand exists. Every board must answer:

1. What does this brand represent?
2. What is the core metaphor?
3. How does the logo express that?
4. How does the system scale across UI, print, image, and detail?
5. Why does the whole thing feel ownable?

Quick Reference

Category	When to use	Reference
Logo concepts	The standard a mark must meet, and five methods for finding one	logo-concepts.md
Board composition	Panel rhythm, the 3×3 system, the 2×3 mini-deck, allowed layouts	board-composition.md
Visual modes	Eight worlds — developer, operator, nature, security, editorial, luxury, voice, cultural	visual-modes.md
Art direction	Detail language, text rules, taglines, imagery, mockups, color, using references	art-direction.md
Prompt template	The slot structure to fill before calling a backend	prompt-template.md

Step 1 — Strategy before pixels

Never generate before inferring the strategy. Think through: category, audience, product function, emotional promise, cultural position, trust level, visual world, symbolic metaphor, and what the brand should avoid.

The visual system must be based on meaning. Symbols are derived, never picked at random:

Category	Core ideas	Symbol logic
Developer tool	building, speed, precision, control	cursor, frame, bolt, scaffold, grid
AI assistant	delegation, intelligence, clarity	spark, orbit, signal, path, node
Security	protection, vigilance, boundary	shield, eye, seal, protected core
Gaming / betting	chance, reward, tension, speed	dice, gem, card, signal, trophy
Voice AI	sound, rhythm, command, flow	waveform, mic, orb, speech path
Compliance	trust, order, rules, protection	seal, dog, badge, document, shield
Drone / robotics	flight, control, vision, mission	wing, owl, crosshair, path, zone
Luxury / editorial	taste, material, ritual, restraint	monogram, seal, paper, emboss, mark
Productivity	focus, momentum, clarity	path, check, block, calendar, light

If the project has a jtbd.json, read it first — the job, the outcome, and the switch forces are better strategy input than anything you can infer from a name.

Step 2 — Compose the board

Default output, unless the user says otherwise: one brand-kit overview image, a 3 × 3 grid, 4:3 or 16:10, clean presentation grid, consistent gutters, minimal text, every panel visibly connected.

Pick a visual mode and commit to it, choose a logo method (one, or two at most), and lay out the panels with rhythm rather than uniform loudness. See the Quick Reference above for all three.

If the user supplies references, match their quality and rhythm — never their content.

Anti-generic rules

Never produce: random floating icons · generic startup gradients · overdesigned logos · meaningless blobs · messy layout collages · fake tiny UI · inconsistent logo marks · too many colors · cheap neon · stock-template brand boards · corporate PowerPoint slides · soulless SaaS dashboards.

When a board is not working, the answer is almost never "more". Make it quieter, sharper, and more intentional. Before shipping, remove one accessory.

Step 3 — Explore directions, then hand off

1. From the brief (or the project's jtbd.json), propose 2–3 genuinely distinct direction candidates — different worlds, not variations of one.
2. Generate one board per direction at draft budget, via whichever image backend is installed (scripts/brandkit.py backends).
3. The user picks a winner. Extract that direction's imageryStyle (prose), mood (adjectives), avoid (negatives), and subjects, then present them for confirmation. Never write without it.
4. On confirm, run the handoff:

# Existing token set: writes $extensions and regenerates DESIGN.md
python3 scripts/brandkit.py handoff --direction direction.json --tokens <set>.tokens.json

# No token set yet: saves brand-block.draft.json
python3 scripts/brandkit.py handoff --direction direction.json --out-dir <project-dir>

A saved draft is imported automatically by humane:design-tokens setup-edit run in the same directory — so the new token set is born with its art direction attached.

Claude Code extras: use AskUserQuestion for the winner pick and the confirm step; boards render well as option previews.

Tests

cd brandkit && python3 -m unittest discover -s tests -v