Conciseness

Fourteen lines with zero padding: every sentence adds decision criteria Claude does not already have ("changes already verified by types, linting, or schema validation", "smallest test level that protects the behavior"). Not score 4 because there are no over-explanations to trim — it never explains what testing is or why it matters.

Actionability

The guidance is concrete and directive for an instruction-only skill: an explicit prioritize list (domain invariants, edge cases, public interfaces, persisted data, security boundaries, integration seams), an explicit skip list, and a decision gate with a defined branch ("If that case is weak, do not add one"). Not score 5 because no worked example or tie-breaking guidance is given for the ambiguous middle cases the gate will actually surface (e.g., a refactored public interface with existing coverage); not score 3 because the categories and the decision test are specific enough to act on directly.

Workflow Clarity

This is a simple single-purpose skill under 50 lines, so the simple-skill exception applies: the single decision action is unambiguous, sequenced (priorities → skip rules → level choice → final gate), and the final gate is an explicit validation checkpoint with a defined failure branch. Not score 4 because the sequence and checkpoint are already fully explicit for the skill's scope; no multi-step workflow is missing.

Progressive Disclosure

Under 50 lines with no external material needed and no bundle files present, so a well-organized single file is the correct structure. Sections are logically ordered (principle → prioritization → skip rules → level selection → final gate) with no monolithic wall of text and no nested references. Not score 4 because there is nothing that belongs in a separate file and no organization gap to point to.

Specificity

Names the domain (test-worthiness decisions) and two concrete actions — "Decide whether a change warrants new tests" and "choose proportionate verification" — but stops there. Not score 4 because it does not list several specific actions or enumerate what the verification choices are (e.g., unit vs integration vs smoke); not score 2 because the two actions it names are genuinely concrete rather than generic.

Completeness

Explicitly answers what ("Decide whether a change warrants new tests and choose proportionate verification") and when ("Use before adding tests, when planning validation, or when reviewing whether a proposed test protects meaningful behavior") with three concrete trigger scenarios. Not score 4 because the when-clause is fully explicit rather than merely present.

Trigger Term Quality

Good natural phrases users would actually say: "adding tests", "planning validation", "reviewing" a proposed test. Not score 5 because common synonyms like "write tests", "unit tests", "test coverage", or "TDD" are absent; not score 3 because the terms present are the natural phrasings, not jargon.

Distinctiveness Conflict Risk

Has a clear niche — deciding *whether* to test, not how to write tests — which distinguishes it from test-writing and general verification skills. Not score 5 because "when planning validation" and "proportionate verification" are broad enough to occasionally overlap with general verification/review skills; not score 3 because the core trigger (test-worthiness judgment) is distinct.