Conciseness

Every line is instruction — scope rules ("Review what the user names: uncommitted or staged changes, a branch, a PR link, or files"), priorities ("Order of importance: correct, safe, holds under load, tested, fast, lean"), and output format. No concepts Claude already knows are explained; brief rationale lines like "This is a report the user asked for, so give it in full" each change behavior, so every token earns its place.

Actionability

Fully executable guidance for an instruction-only skill: exact output openers ("Start with `What this change does:`"), a worked four-part finding example with a real fix (`datetime.now(ZoneInfo("Europe/Berlin")).date()`), exact verdict formats ("Verdict: Ship."), the lean summary line, the not-checked line, and the empty case ("Looks good. Ship."). The output is copy-paste ready and covers the common cases.

Workflow Clarity

Clear four-phase sequence (Understand first → Look for → Check before you report → Output) with an explicit validation gate before reporting: "Every finding needs a concrete case ... No case, no finding" and "Re-read the lines and confirm: the caller exists, the value can really be empty, the code really is unused." This is a read-only review skill, so the destructive/batch validation cap does not apply; the checklist matches the 5 anchor.

Progressive Disclosure

No bundle files exist (no references/, scripts/, or assets/), and the ~90-line body is a cohesive single-task workflow in four numbered sections with nothing that clearly belongs in a separate file. Well-organized and easy to navigate; splitting it would hurt rather than help, so the structure matches the top anchor.

Specificity

Lists six concrete review capabilities ("is the logic right, is it safe, does it hold under real load, is risky code tested, is it fast enough, and is every line needed") plus concrete behaviors ("Reads the connected code, not only the diff", "Each finding is explained in plain English"), covering the skill's full scope. This matches the 5 anchor (multiple specific concrete actions, comprehensive coverage); a 4 would require evident gaps, and none are present.

Completeness

Explicitly answers both questions: what ("Quality review of a change" with the six named dimensions, reading connected code, plain-English findings) and when ("Use for 'review this', 'code review', 'review the last commit', 'review my PR', 'is this over-engineered', /ponytail-review") with concrete trigger phrases. This is the 5 anchor verbatim in structure.

Trigger Term Quality

Natural phrases users would actually say: "review this", "code review", "review the last commit", "review my PR", "is this over-engineered", plus the /ponytail-review slash command. Coverage spans change/commit/PR synonyms; no meaningful natural terms are missing for this domain, so it fits the 5 anchor rather than 4's 'a few natural terms missing'.

Distinctiveness Conflict Risk

Triggers like "code review" and "review my PR" carry minor overlap risk with other review-oriented skills, but the distinctive scope (load, tests, lean) and the "is this over-engineered" trigger give it a mostly-distinct niche. Best fit is 4 (mostly distinct; minor overlap risk), not 5's minimal conflict risk.