Conciseness

The body is dense and contract-style with no explanation of concepts Claude already knows, but the two Focus ID tables (~40 lines) duplicate the canonical tables already in the framework references, and some boundary rules are restated across sections (e.g. 账号运营子模块不进 template appears twice). These are minor instances that could be trimmed, matching anchor 4 rather than the fully lean anchor 5.

Actionability

Fully executable, copy-paste-ready CLI invocations are given for all three subcommands (report, build, update) with real flags for both video and note kinds, plus concrete storage paths, a worked user-input translation example, and explicit edge-case rules (「口播密度标为未提供…不能用 0 代替缺项」). The examples cover the common cases completely.

Workflow Clarity

The report → build → update pipeline is clearly sequenced with numbered agent checklists and explicit validation checkpoints (kind mixing "混型直接报错", "--focus 按作品类型校验", coverage marking "少量样本不得称为稳定结论", conflict preservation for user input). Not 5 because there is no explicit output-verification or error-recovery loop after build/update; the validation that exists is mostly precondition checking rather than feedback loops.

Progressive Disclosure

The body is a clear operational overview that appropriately splits dimension frameworks into two one-level-deep reference files (references/video-dna-framework.md, references/note-dna-framework.md — both verified to exist), signaled both in the --kind table and in the 参考资料 section with one-line content descriptions, making navigation easy.

Specificity

The description lists several concrete actions ("单篇作品（视频 / 图文）生成 report", "按 dna-id 聚合…", "推导 main agent 的 Brief / 图文生产 template") with a detailed enumeration of aggregation dimensions, but the headline action "提取抖音作品 DNA" is metaphorical jargon and capabilities like update and input conventions are uncovered — minor gaps rather than comprehensive coverage.

Completeness

The "what" is clear and detailed (generate report, aggregate by dna-id across enumerated dimensions, derive Brief/template), but there is no "Use when…" clause or equivalent explicit trigger guidance, which caps completeness at 3 per the rubric guideline. Not 2 because the "what" is explicit rather than vague.

Trigger Term Quality

Good keyword coverage with natural domain terms ("抖音", "视频 / 图文", "选题", "标题与封面", "口播文案", "业务植入套路", "CTA", "Brief"), but common user phrasings such as 风格拆解, 对标账号, or 爆款 are missing. Not 5 because a few natural synonyms users would actually say are absent; not 3 because coverage is broad and domain-appropriate.

Distinctiveness Conflict Risk

Clear niche with distinct triggers: "抖音作品 DNA", "dna-id", "口播文案子模块", "图文生产 template" are highly specific to this Douyin style-profiling tool, with minimal overlap risk against any generic skill.