The body is dense and information-dense, assumes Claude's competence (no explanations of what Node/FFmpeg/contrast are), and routes detail to references; it stays lean rather than padded with concepts Claude already knows.

Provides fully executable, copy-paste-ready commands and pipelines throughout ('npx hyperframes lint', 'jq -e ".ok"', '[ -s "$OUTPUT" ]', ffprobe checks), not vague direction or pseudocode.

The Scaffold→Write→Lint→Validate→Inspect→Preview→Render sequence is explicit with validation checkpoints ('Run lint, validate, and inspect before preview', post-render verification, --strict gating), giving clear feedback loops rather than implicit steps.

A concise overview with a Routing table pointing to six one-level-deep references (all verified present in references/), with content appropriately split and easy to navigate; no nested/deep reference chains.

Enumerates many concrete CLI actions ('init, add, catalog, capture, lint, validate, inspect, layout, snapshot, preview, play, render, publish, lambda, doctor, browser...'), matching the 'lists multiple specific concrete actions' anchor rather than the level-2 'some actions' anchor.

Explicit 'what' ('HyperFrames CLI dev loop') and explicit 'when' ('Use when running npx hyperframes... or when troubleshooting the HyperFrames build/render environment'), satisfying both halves; not level 2 because the 'when' is explicit, not implied.

Trigger terms are the natural subcommand names users would say ('npx hyperframes init', 'lint', 'render', 'tts', 'remove-background') plus 'troubleshooting the HyperFrames build/render environment' — full coverage of natural terms rather than jargon-only.

A clearly distinct niche (HyperFrames video build/render CLI) with command-specific triggers unlikely to fire for unrelated skills; not level 2 because the triggers are product-specific rather than generic.