The body is telegraphic and information-dense (bold labels, · separators, fragments) and never explains concepts Claude already knows (UX, IA, accessibility); it assumes competence and packs orchestration doctrine tightly, matching the lean/efficient anchor rather than the padded anchor 2 example that adds unnecessary justification.

Concrete, executable commands are given verbatim (resolve_customization.py, memlog.py init/append with exact flags, python3 webbrowser openers) plus specific file destinations (.working/, imports/, mockups/, wireframes/) and locked section orders; the {placeholder} variables are a documented template mechanism resolved at activation, not pseudocode, so guidance is copy-paste-ready post-resolution.

Multi-step processes are explicitly sequenced (numbered Activation steps, ordered Finalize outcomes, Create/Update/Validate mode flows) with validation checkpoints and feedback loops (Reviewer Gate, rubric-walker Pass 1 coverage checks, mock-coverage walk, activation-completion confirmation, "resolve findings before polish" gate); this matches the clear-sequence-with-explicit-validation anchor.

The body is an orchestration overview that points one level deep to real, well-signaled references (references/design-md-spec.md, headless.md, validate.md, creative-tools.md) and example assets (design-example-*, experience-example-*), with detailed specs appropriately externalized and no deep nesting; all cited bundle paths exist.

"Plan UX patterns and design specifications" names the UX domain and a single action verb ("Plan") applied to two objects, matching the anchor that names a domain and some actions but is not comprehensive; it falls short of score 3 because it does not list multiple distinct concrete actions (cf. "extract, fill, merge").

It clearly answers both what ("Plan UX patterns and design specifications") and when ("Use when the user says ..." with three explicit triggers), satisfying the anchor for an explicit Use-when clause; it is not 2 because the when is explicit, not merely implied.

Three explicit, natural phrasings a user would actually say ("lets create UX design", "create UX specifications", "help me plan the UX") give good coverage of natural terms; it is above score 2 because the triggers are genuine user-voiced phrases rather than one generic label.

The UX-qualified niche ("UX patterns", "UX specifications", "plan the UX") with distinct UX-only triggers is unlikely to fire for adjacent skills; it is above 2 because the triggers are narrowly scoped to UX rather than a generic "design" or "documents" surface.