Lean body with no exposition of concepts Claude already knows; the format spec, three depth-differentiated examples, and tips each earn their place by demonstrating the exact annotation format the skill teaches.

Provides a concrete copy-paste format template, three fully worked examples, explicit symbols (→ ✓ ✗ ←), and depth-selection guidance — giving specific, actionable instruction rather than vague direction.

A sequenced format template and case-based depth guidance are present, but the actual procedure for producing a trace (identify stages → pick values → annotate → verify) is implicit rather than given as an ordered, checklisted workflow.

A single self-contained file organized into clearly headed sections (Format, When to use which depth, Examples, Tips) with no nested references; for this format-demonstration skill the inline examples are appropriately integral rather than bloating a separate file.

Lists multiple concrete actions — 'Trace and document how data transforms', 'showing intermediate state at each step with concrete example values' — rather than vague language, matching the anchor for multiple specific concrete actions.

Explicitly answers both what ('Trace and document how data transforms...') and when via an explicit 'Use when...' clause enumerating triggering situations, so it is not capped at 2.

Covers natural terms users would actually say — 'data pipeline', 'complicated codepaths', 'how does X get to Y', 'dataflow walkthrough', 'code review', 'design doc' — giving good coverage of natural phrasings.

It carves a clear niche (annotated dataflow tracing through pipeline/function chains) with distinct triggers unlikely to fire for unrelated skills; not generic enough to conflict broadly.