Conciseness

The body is overwhelmingly lean, imperative guidance — tables for claim categories and figure plans, an exact problem-statement format, a concrete dependency-map sketch, and a full output template — with no explanations of concepts Claude already knows. Minor trimmable instances exist: the 9-line reviewer-routing blockquote at the top, the nested "Resolve this reviewer slot first" blockquote in Step 6 that repeats it, and the dangling "Adapted from the refinement pattern in /research-refine" line. This matches anchor 4 ("Efficient; minor instances of over-explanation that could be trimmed") rather than anchor 5, and is above anchor 3 because the unnecessary material is a small fraction of otherwise tight content.

Actionability

For an instruction-only skill the guidance is fully executable: an exact problem-statement template ("The technical problem to be solved is how to [specific technical objective] given [specific technical constraint]"), a concrete novelty test for core features ("if you remove this feature, the invention is no longer novel"), two lookup tables, a worked claim-hierarchy example, a complete copy-paste reviewer spawn_agent spec with model/effort/message, and a complete output markdown template. This matches anchor 5 ("Fully executable; copy-paste ready... specific examples cover the common cases"); the scoring note for instruction-only skills confirms absence of runnable code is not penalized when guidance is this actionable.

Workflow Clarity

Seven clearly sequenced steps end in an explicit validation checkpoint (Step 6 fresh-agent review with five concrete assessment questions) plus a stop-and-ask rule if delegation is unavailable — matching "clear sequence with most checkpoints present". It falls short of anchor 5 because the feedback loop is not spelled out: MAX_REFINEMENT_ROUNDS = 3 is declared but the revise-then-revalidate iteration (how reviewer findings are applied and when rounds stop) is left implicit. There are no destructive or batch operations, so the workflow-clarity cap of 3 does not apply.

Progressive Disclosure

Structure is good: the body is sectioned into Constants/Inputs/Workflow/Key Rules, and external detail is pushed to clearly signaled one-level-deep references — "Load ../shared-references/patent-writing-principles.md for the Problem-Solution-Advantage framework" and the reviewer-backends pointer for the reviewer slot. Minor gaps keep it at anchor 4 rather than 5: the reviewer-backends reference is buried inside a dense blockquote at the top rather than a clean navigation section, and no bundle-local references/ or scripts/ files exist to carry the tables or spawn spec. It is well above anchor 3 since references are present, signaled, and only one level deep.

Specificity

The description names the domain and one-to-two concrete actions — "Structure a raw invention idea into a formal invention disclosure" and "formalize a rough idea into a patent-ready structure" — but these are two phrasings of the same single action. It does not mention the concrete sub-capabilities the skill actually performs (feature decomposition, claim-category planning, drawing plan, dependency mapping, reviewer validation), so coverage is not comprehensive. It fits anchor 3 ("Names domain and 1-2 concrete actions, but not comprehensive") better than anchor 4, which expects several distinct specific actions listed; it is clearly above anchor 2 because the action named is specific rather than generic.

Completeness

It clearly answers both questions: the "what" is "Structure a raw invention idea into a formal invention disclosure", and the "when" is explicit via "Use when user says \"构建发明\", \"structure invention\", \"发明构建\", \"invention disclosure\", or wants to formalize a rough idea into a patent-ready structure." This matches anchor 5 exactly — both what and when with concrete trigger phrases — and exceeds anchor 4, where the 'when' is present but could be more specific.

Trigger Term Quality

The description provides comprehensive natural triggers with synonyms: "构建发明", "structure invention", "发明构建", "invention disclosure", plus the intent-based phrase "or wants to formalize a rough idea into a patent-ready structure". Bilingual synonym coverage plus a natural user-intent phrasing matches the comprehensive-synonyms anchor; it is above anchor 4 ("a few natural terms missing") because no obviously common phrasing of this niche is omitted.

Distinctiveness Conflict Risk

The niche is unambiguous (patent/invention disclosure structuring) and the triggers — including the Chinese phrases "构建发明"/"发明构建" — are distinct phrases users would only say when they want this specific skill. Conflict risk with other skills is minimal, matching anchor 5's "clear niche with distinct triggers"; it is well above anchor 4's "minor overlap risk with closely related skills" since no sibling skill would plausibly claim these exact triggers.