Lean and dense — uses tables, bulleted field schemas, and 'rejects/risk' framing rather than prose, assuming Claude's competence and never padding with concepts Claude already knows.

Gives concrete executable guidance — exact counts ('Generate 3 to 5 territories,' '8 to 15 candidates'), a six-criteria naming filter, named narrative shapes, and a fillable output template — copy-ready direction with no vague abstraction.

An explicit 8-step workflow sequences divergence-then-convergence across four stages with gating checkpoints (e.g. 'A short-listable name passes all six'); ideation is non-destructive so the missing error-retry loop does not cap the score.

Body is a concise overview that points to two clearly-signaled one-level-deep references (ideation-output-template.md, naming-evaluation-rubric.md), both real files in references/, keeping detail out of SKILL.md.

Lists multiple concrete actions — 'Generate, evaluate, and narrow brand concepts' across 'positioning territories, naming candidates, mood directions, and narrative angles' — matching the highest anchor for specific concrete actions.

Explicitly answers both what ('Generate, evaluate, and narrow brand concepts...') and when ('Use this skill whenever the user is in the early phase of brand creation...'), with an explicit trigger list.

Provides broad natural-term coverage a user would actually say — 'brand ideation,' 'naming,' 'brand name,' 'mood board,' 'exploring brands,' 'brand brainstorm' — with no jargon-only gaps.

Clear niche as the early divergent-convergent ideation phase, positioned explicitly upstream of brand-identity/style-guide/voice, with distinct triggers unlikely to fire for sibling skills.