Conciseness

The body is mostly directive with little concept explanation, but it could be tightened: the EXTREMELY-IMPORTANT block repeats the same rule three times ("YOU ABSOLUTELY MUST invoke the skill", "YOU DO NOT HAVE A CHOICE", "This is not negotiable"), and several Red Flags rows overlap (e.g., 'This is just a simple question' vs 'This doesn't count as a task'). Not score 2 because it never explains concepts Claude already knows; not score 4 because the redundant emphasis and overlapping table rows are trimmable without losing meaning.

Actionability

The guidance is concrete and executable for an instruction-only skill: an exact announcement pattern ("Using [skill] to [purpose]"), a checklist rule ("If it has a checklist, create a todo per item"), and worked priority examples ("'Let's build X' → superpowers:brainstorming first"). It falls short of score 5 because the body never explains HOW to find available skills — the core capability the description promises — leaving a minor gap.

Workflow Clarity

The single workflow is clearly sequenced: check for skills BEFORE any response or action → invoke matching skill → announce → follow exactly (todos per checklist item), with priority ordering and a plan-mode precondition. Not score 5 because there is no checkpoint or fallback guidance (e.g., what to do when no skill clearly applies, or when a invoked skill turns out wrong mid-use beyond the single sentence); not score 3 because the sequence present is coherent and unambiguous.

Progressive Disclosure

Platform-specific detail is correctly split into one-level-deep, clearly signaled references (references/codex-tools.md, pi-tools.md, antigravity-tools.md, hermes-tools.md — all verified to exist). Not score 5 because references/gemini-tools.md exists in the bundle but Gemini is missing from the platform list, leaving a real file unnavigable from the skill.

Specificity

The description names its domain ("how to find and use skills") and one to two concrete actions ("requiring skill invocation before ANY response including clarifying questions"), but it does not comprehensively list what the skill actually covers (skill priority, red flags, platform adaptation). It is above score 2 because the actions stated are concrete, not generic; it is below score 4 because coverage stops at two actions.

Completeness

Both parts are explicit: the 'when' is a concrete trigger clause ("Use when starting any conversation") and the 'what' is stated directly ("establishes how to find and use skills, requiring skill invocation before ANY response including clarifying questions"). Not score 4 because the 'when' is not merely adequate — it is a specific, unambiguous trigger phrase.

Trigger Term Quality

Natural trigger phrases are present: "Use when starting any conversation", "skill invocation", "clarifying questions" — phrasing a user or harness would plausibly match on. It falls short of score 5 because there are no synonyms or variations (e.g., 'beginning a session', 'skill check', 'before answering') to broaden matching.

Distinctiveness Conflict Risk

The trigger "starting any conversation" fires alongside virtually every other skill, creating maximum overlap risk — a deliberate design for a meta-skill, but per the rubric it is very broad with high conflict risk. Not score 1 because the topic itself (skill invocation discipline) is a recognizable niche rather than generic filler; not score 3 because the overlap is with all skills, not just closely related ones.