Conciseness

No concept explanations or padding; every line is operational detail Claude cannot know elsewhere — tool parameters ('brand_read sections: ["kit"], detail: "full"'), precedence order ('owner > owner's documents > owner's earlier rules > site > research'), triage budgets, and time-pressure fallbacks. Lean and efficient; every token earns its place.

Actionability

Fully concrete for an instruction-only skill: exact tool calls with parameters ('gooseworks:knowledge_import with url', 'brand_update patch.kit'), exact file paths ('brand/records/<YYYY-MM-DD>-v<N>.md, imported as a brand document'), a verbatim marker string to write, named sub-playbooks to run (atoms/claims-audit, atoms/brain-write-back), and copy-ready owner messages ('Connect GooseWorks first — I learn and save your brand through it.').

Workflow Clarity

Eight numbered steps with mode dispatch up front (full / resume / update / one claim / client sign-off back / sweep), save checkpoints after steps 3–5, and an explicit validation gate in step 8: 'Re-read fresh, not from this conversation … If a probe needs this conversation, fix the save', plus error handling ('Report anything that did not save'). Clear sequence with explicit validation steps and feedback loops for error recovery.

Progressive Disclosure

The body is a genuine overview pushing detail to one-level-deep, conditionally-loaded references ('if regulated, load references/category-rules.md') and all five referenced references/ files exist in the bundle. However, references/brand-record-template.md is present but never linked from the body (its content is re-described inline in 'Output and write-back'), and the referenced paths atoms/claims-audit, atoms/voice-calibrate, atoms/brain-write-back and tests/test-by-hand.md do not exist in the provided bundle — navigation gaps that keep this below the anchor-5 'easy navigation' bar.

Specificity

Names around six concrete actions ('audits every claim (including ones research marked approved) and gets the ones that matter signed off with proof, scope and review date', 'turns the voice into rules the owner confirms', 'checks logo, colours and fonts') — anchor-5 coverage. However, the first-person opener 'Learn my brand once and stop asking me every time' violates the third-person voice rule ('Learn my brand', 'stop asking me'), which the rubric penalizes by reducing specificity by 1.

Completeness

The 'what' is explicit (read existing research, ask only gaps, audit claims, confirm voice rules, check kit, save continuously) and the 'when' is an explicit 'Use when someone says …' clause with concrete trigger phrases, plus a 'Not for …' exclusion list. This matches the anchor-5 example pattern exactly.

Trigger Term Quality

Twelve natural user phrases — 'learn my brand', 'set up my brand', "here's our brand book", 'remember how we talk', 'what can we claim', 'can we say this', 'approve our claims', 'sign off our voice', 'our price or product changed', 'the client approved the claims', 'pick up where we left off', 'check our claims are still current' — exceed the anchor-5 example's synonym coverage; no common variation is missing.

Distinctiveness Conflict Risk

Clear niche — owner sign-off of brand claims and voice within GooseWorks — with distinct triggers and explicit boundary guidance: 'Not for who our customers are and what they say (use research-audience) or writing positioning from scratch (not covered yet)'. This routes away sibling research skills, minimizing conflict risk.