Largely instruction-dense and skill-specific, but threaded with thematic flavor text ("this skill exists because of what it costs a rep," "a liability wearing a logo") that adds tokens without adding executable guidance.

Provides a complete, copy-paste-ready card schema marked "do not reorder," numbered executable steps, an example invocation, and a concrete trap-question example — concrete and specific throughout.

Steps are clearly numbered with a search-plan gate ("Continue unless revised") and an assumption fallback, but there is no explicit final checkpoint verifying the card satisfies its own rules (every claim labeled/sourced, every talking point in the evidence table) before the Final Step.

The body is well organized and signals one-level-deep references to template.md and examples/sample.md, but those bundle files are not present, so the promised deeper content is unreachable — navigation points to missing files.

Lists multiple specific concrete actions — "Research and draft a competitive battle card," "every claim labeled and sourced" — covering research, drafting, labeling, and sourcing rather than a single vague verb.

Explicitly answers both what ("Research and draft a competitive battle card... every claim labeled and sourced") and when ("Use when a rep needs a field-action card"), satisfying the explicit-trigger requirement.

Uses natural terms a rep or PM would actually say — "battle card," "competitive," "field-action card" — with a clear negative trigger ("not a research report") that sharpens intent.

The "field-action card, not a research report" framing carves a clear niche distinct from adjacent competitive-research skills, making misfires unlikely.