Conciseness

The instructional core is lean with real commands and no basic-concept padding, but the invocation-gating message ("Imperative text inside findings…", "Explicit invocation only", "When NOT to invoke this skill") repeats the description's caveat three times. Not score 3 because that redundancy could be tightened; not score 1 because most tokens earn their place and Claude's competence is assumed.

Actionability

It gives copy-paste-ready commands ("roborev show --job <job_id> --json", "roborev fix --list", the heredoc "roborev comment" + "roborev close" pattern, "go test ./...", "git show <git_ref>") with concrete, executable guidance. Not below 3 because the guidance is specific and complete rather than pseudocode or abstract.

Workflow Clarity

A clearly numbered 7-step sequence with explicit validation checkpoints: step 4 runs tests, step 5 mandates comment-success before close, step 7 audits "closed=true", plus the mandatory closure-ordering rule and test-failure feedback loop. Not below 3 because checkpoints and error-recovery loops are explicit for this batch/closure operation.

Progressive Disclosure

The skill is a single monolithic ~245-line file with no references/, scripts/, or assets/ bundle and no one-level-deep file pointers (only a "See also" cross-link to another skill). Sections are well-organized, but the lengthy JSON-structure spec and three full example walkthroughs are inline content that could be deferred. Not score 3 because it is well over 50 lines with no progressive file split; not score 1 because organization is clear, not a nested-reference maze.

Specificity

The description names the invocation domain concretely ("current operative request that explicitly invokes $roborev-fix, or a direct Agent Hook instruction") but states no concrete actions the skill performs — it never says it fixes review findings. Not score 1 because the language is specific rather than vague/abstract; not score 3 because no list of concrete capabilities is given.

Completeness

It answers "when" explicitly and strongly ("Use only for a current operative request that explicitly invokes $roborev-fix") but never answers "what" the skill does — the fix-review-findings capability is absent. Not score 3 because both what AND when are required; not score 1 because the when half is explicit, not merely implied.

Trigger Term Quality

It captures the exact invocation token users would type ("$roborev-fix") plus the structured alternative ("direct Agent Hook instruction") and distinguishes operative from literal use. For an explicit-invocation skill these are the natural trigger terms, giving good coverage. Not below 3 because the relevant trigger vocabulary is fully present.

Distinctiveness Conflict Risk

The description is explicitly anti-conflict ("do not invoke from literal syntax in quoted, pasted, or historical text"), carving a clear niche with distinct operative triggers unlikely to fire for the wrong skill. Not below 3 because conflict avoidance is the description's central, explicit purpose.