Conciseness

The body is dense and assumes competence — it jumps straight to commands and file paths with no library-primer padding, and every section carries skill-specific knowledge (version-matched sourcing, corpus install, override ladder). Not a 5 due to real redundancy: tool availability is stated twice ("The same tool is available in the headless `pnpm agent` loop and every built-in app agent" vs. "The headless `pnpm agent` loop and built-in app agent also expose... tools") and "From a generated app directory" appears twice with overlapping reader listings.

Actionability

Fully executable, copy-paste-ready commands throughout: `pnpm action framework-search --pattern "defineAction"`, `docs-search --slug <slug>`, the corpus install command with the version pin, `rg -n ... node_modules/...`, and a `cp` of a named template file to a named destination. The commands cover the common cases (docs lookup, source grep, template reuse, fallback when the action runner is missing), matching the top anchor.

Workflow Clarity

There is a clear decision sequence — unified `framework-search` first, then focused readers, with explicit fallback branches ("If source-search reports that template source is unavailable, install the matching corpus package...", "If the action runner is unavailable, search the package directly") and an instruction to refine rather than accept truncated results. Not a 5 because the sequencing is implicit across sections rather than laid out as an ordered flow, and the section ordering interleaves fallback and primary paths; not a 3 because checkpoints and error-recovery branches are explicitly present.

Progressive Disclosure

No bundle files exist (references/, scripts/, assets/ are absent), and the body is a self-contained, well-sectioned ~150-line document with clear headers (Rule, Why, How, Useful Slugs, Don't) — good structure with appropriate placement of most content. Not a 5 because some content is inline that could plausibly live in reference files (the "Authoring visual docs blocks" section and the "Useful Slugs" table are peripheral to the lookup workflow), and the repeated reader/tool listings blur the navigation slightly.

Specificity

Concrete actions are named — "find version-matched Agent-Native framework docs and source bundled in node_modules" — and the target surface is enumerated ("@agent-native/core APIs, generated apps, workspaces, templates, or advanced features"), matching the 'several specific actions; minor gaps' anchor. Not a 5 because the only verbs are 'find'/'answer questions'; it doesn't enumerate the concrete capabilities (search docs, search source, copy template patterns) that the skill actually delivers.

Completeness

Both questions are answered explicitly: what ("How to find version-matched Agent-Native framework docs and source bundled in node_modules") and when ("Use before implementing or answering questions about...") with concrete trigger phrases. Not a 4 because the 'when' clause is already explicit and trigger-specific, not merely present-with-room-for-improvement.

Trigger Term Quality

Natural trigger phrases are present — "implementing or answering questions about @agent-native/core APIs", "generated apps", "workspaces", "templates", "advanced features" — which a user of this framework would plausibly say. Missing a few natural variations (e.g., "look up framework docs", "Agent-Native", "how does X work"), so it fits the 'good keyword coverage; a few natural terms missing' anchor rather than the comprehensive 5.

Distinctiveness Conflict Risk

It names a distinct niche — version-matched @agent-native/core docs bundled in node_modules — with trigger terms specific to that framework, giving it a clear niche with minimal conflict risk against generic documentation or web-search skills. It would not plausibly fire for unrelated skills.