Conciseness

The body is lean and assumes competence — terse condition→action bullets with no explanation of concepts Claude already knows ("If you change a function signature... update the corresponding tests before committing"). It is not a 5 because the "✗ Incorrect" examples and the "Anti-Patterns" section restate the same violations ("committed without updating casting.test.ts" vs "Committing API changes without test updates"), which could be consolidated.

Actionability

Concrete, actionable guidance throughout: named artifacts ("EXPECTED_FEATURES", "distributed-mesh.md", "auth.test.ts"), explicit sequencing ("in the same commit"), and a diagnostic rule ("verify test assertion arrays match filesystem state"). It is not a 5 because it never tells how to locate the assertion arrays (e.g., grep for EXPECTED_) — the one gap between instruction and execution.

Workflow Clarity

This is a simple single-purpose skill (~30 lines) whose single action — sync tests/assertions in the same commit — is unambiguous, so the simple-skill exception applies. It also includes a checkpoint pattern ("CI failures → check assertions first") that orders diagnosis before debugging; nothing is incoherent or missing for the skill's scope.

Progressive Disclosure

The skill is under 50 lines with no need for external references (no references/, scripts/, or assets/ bundle exists, and no body link is broken), and its four clearly-labeled sections (Context, Patterns, Examples, Anti-Patterns) are well organized — the rubric's stated condition for a 5 in this case.

Specificity

"Update tests" is one concrete action anchored to the domain "when changing APIs", matching the anchor for naming a domain plus 1-2 concrete actions. It is not a 4 because it lists only a single action and omits the skill's other capabilities (assertion-array sync, CI triage).

Completeness

Both what ("Update tests") and when ("when changing APIs") are explicitly stated in the single clause, so the missing-trigger cap at 3 does not apply. It falls short of 5 because the when-clause covers only one trigger and gives no concrete trigger phrases for the assertion-sync and CI-failure cases the body covers.

Trigger Term Quality

"tests" and "changing APIs" are natural terms a user might say, but common variations users actually use — "CI", "test suite", "assertions", "coverage", "failing tests" — are absent. This matches "some relevant keywords but missing common variations or synonyms" rather than the good-coverage anchor.

Distinctiveness Conflict Risk

"Update tests when changing APIs" carves a clear niche (test hygiene tied to interface changes) that is mostly distinct from generic skills. It is not a 5 because a general testing or CI skill could plausibly trigger on the same phrasing.