agentsclimarketplace
Sign in

Brand

Skill event4u-app/agent-config/dist/agent-src/skills/brand

Grounded brand decisions from a curated corpus — archetype, voice, naming, colour psychology, logo-style fit, messaging frameworks, archetype→type mapping. Use to ground brand strategy and identity.From its SKILL.md

Install
npx -y skills add event4u-app/agent-config --skill brand

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • 7 stars7 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
  • runs commandsInstructs the agent to run 1 command, including `./scripts-run <skills-root>/corpus-grounding/scripts/ground search --manifest <skills-root>/brand/data/manifest.json "<brand brief: sector + intent + audience>" [--domain archetype|voice|naming|color|`.

SKILL.md

5.2 KB, ~1.2k tokens by cl100k_base, as published. Nobody here has run it

brand

The grounded source for brand decisions — a second instance of the ADR-061 corpus-grounding layer (corpus-grounding), after design-intelligence. Branding is the layer that constrains UI: the corpus grounds brand strategy and identity decisions (archetype, voice, naming, colour, logo style, messaging, archetype→type filter) as a constraint set the human confirms — never the final brand. No forked engine; this plugs into the shared one via a manifest.

Corpus: 7 tabular CSVs under data/ — 12 brand archetypes, a voice-and-tone matrix, naming patterns, color psychology by industry, logo-style ↔ industry fit, messaging frameworks, and typography-principles (archetype → pairing-filter Grounding, the layer that upgrades typography-system stage-2). Provenance: ATTRIBUTION.md; manifest: data/manifest.json.

When to use

  • A brand decision needs grounding: which archetype fits, what voice/tone, how to name, which colour direction, which logo style, which messaging framework, or which type pairing-filter an archetype implies.
  • Before brand-strategy / brand-identity commit to a direction — those skills consult this corpus first.
  • When typography-system needs the brand-aware (archetype → pairing-filter) upgrade — query the typography domain here.

Procedure: consult the brand corpus

  1. Ground or search (paths resolve skill-relative; works from any cwd):

    ./scripts-run <skills-root>/corpus-grounding/scripts/ground search \
  --manifest <skills-root>/brand/data/manifest.json \
  "<brand brief: sector + intent + audience>" \
  [--domain archetype|voice|naming|color|logo|messaging|typography] \
  [--filter "Archetype Fit=Ruler"] [--json]

    <skills-root> is ~/.claude/skills/ for Claude Code installs, src/skills/ inside this repo.

  2. Read confidence + every evidence_gap line before trusting any row — surface them; the human signs off on what the corpus could NOT support.

  3. Propose grounded options (archetype + voice + colour + logo + messaging), each cited per corpus row, with alternatives — the human confirms.

  4. The confirmed selections become the brand token + voice constraint set that brand-to-tokens, brand-consistency, and pack-ai-image's brand-asset generation consume.

Output format

  1. Grounded brand candidates per domain (archetype, voice, naming, colour, logo style, messaging, type filter) — each selection cited per corpus row.
  2. The grounded output's confidence label + every evidence_gap line, verbatim.
  3. Alternatives per domain so the human can swap before confirming.

Do NOT

  • Do NOT let the corpus author final brand copy or names — it supplies constraint sets and patterns; final strings are agent-written and human-confirmed.
  • Do NOT fork the engine — plug in via the manifest (ADR-061 §2).
  • Do NOT override the consumer's existing brand tokens / voice profile with corpus rows — consumer brand is the source of truth; the corpus fills gaps (see brand-source-of-truth).
  • Do NOT hide low confidence — the user signs off on the gaps too.

Gotchas

  • Corpus grounds pre-action selection — not mid-task reference and not output validation (brand-consistency owns validation).
  • Keep queries brand-shaped ("luxury law firm rebrand", "playful kids snack brand") — generic words land on the default archetype domain.
  • An empty result is a legitimate outcome — surface the evidence gap and proceed on priors; never widen filters to force a hit.
  • BM25 drops tokens ≤2 chars — pad short queries ("B2B", "AI") with companions.

See also

Policies

  • Provenance: ATTRIBUTION.md — original-authored corpus from public brand frameworks; shared engine attributed in design-intelligence/ATTRIBUTION.md.
  • Refresh: quarterly per the manifest; re-review archetype↔type rows when the font-pairings Reference refreshes.

What ships with it: 10 files

42.3 KB alongside SKILL.md

data/

evals/

Gives 0 of the 12 instructions most marketing audience skills give in ~1.2k tokens

Counted across 690 of the 894 authors here whose files we hold, read 2026-08-07

  • Apply Poppins font to headingsin 41 of 690, across 6 files
  • Apply Lora font to body textin 41 of 690, across 6 files
  • Use Arial fallback for headingsin 39 of 690, across 4 files
  • Use Georgia fallback for body textin 39 of 690, across 4 files
  • Maintain text hierarchy and formattingin 39 of 690, across 4 files
  • Use accent colors for non-text shapesin 38 of 690, across 3 files
  • Use RGB values for precise color matchingin 38 of 690, across 3 files
  • Use brand colors for primary text and backgroundsin 36 of 690, across 1 file
  • Read product marketing context file before asking questions, starting, or auditingin 35 of 690, across 23 files
  • Use active voice instead of passive voicein 26 of 690, across 10 files
  • Implement or generate appropriate JSON-LD structured datain 24 of 690, across 17 files
  • Prioritize clarity over clevernessin 22 of 690, across 8 files

Said here and by no other author read

  • search the brand corpus before committing to a direction
  • read confidence and every evidence_gap before trusting a row
  • surface low confidence and evidence gaps to the human
  • propose grounded brand options cited per corpus row
  • provide alternatives per domain for human confirmation
  • use confirmed selections as the brand token set

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.