Conciseness

The body is dense with domain-specific constraints rather than concepts Claude already knows, but the same rule — 「Never 由本 skill 直接执行 ontology CLI，委托 ontology-core」 — is restated at least five times (intro, 调用方式, 子技能依赖 table, 委托命令形态 section, and again in 与 smart-data-analysis 的关系), and the no-token/gateway note appears twice. It fits the 'mostly efficient but could be tightened' anchor rather than the 'minor instances' of the 4 anchor.

Actionability

Concrete command forms are given (`ontology --user-id <accountId> bkn object-type query <kn_id> <ot-id> '<condition-json>' [--limit n]`, `bkn object-type list <kn-id>`) plus explicit search/condition/limit rules and a full worked example deferred to references/tool-examples.md with a real condition-json. It is not a 5 because the body itself only carries placeholder templates; fully copy-paste-ready invocations require the reader to assemble pieces from the reference files.

Workflow Clarity

The sequence is fixed and explicit (metadata KN query → duty KN query derived from step-1 clues → summary), with a copyable progress checklist, and explicit error-recovery/feedback paths: 空结果 → 「放宽 search / 换 KN / 二次澄清」, 命中过少 → rewrite search or loosen condition, 过宽 → lower limit, no clues → 简要反问, missing duty_kn_id → skip and note in summary. This matches the 5 anchor (checklist + feedback loops for recovery) and clearly exceeds the 4 anchor's 'minor validation gaps'.

Progressive Disclosure

The body is a true overview: a per-step reference table clearly signals [references/metadata-search.md], [references/duty-search.md], and [references/tool-examples.md] (all real files), config.json is pointed to for the pipeline/runtime contract, and the references are one level deep with no nested indirection (verified: they only cross-link sibling skills). This matches the 5 anchor's well-signaled, one-level-deep structure.

Specificity

The description enumerates the full pipeline concretely: 「在元数据型知识网络下用 ontology bkn object-type query 检索表/视图实例」「检索相关部门职责与治理边界」「汇总为中文结论（候选表 + 职责要点 + 下一步）」 — multiple specific actions with comprehensive coverage of what the skill does. It even states the execution boundary (CLI delegated to ontology-core), so it is not below the 5 anchor's comprehensiveness.

Completeness

Both halves are explicit: the 'what' is the detailed 3-step pipeline (metadata KN retrieval → duty KN retrieval → Chinese summary), and the 'when' is the explicit clause 「当用户问『表在哪、哪个视图、数据资产归属、谁负责这类数据』时使用」 with concrete trigger phrases. It clearly matches the 5 anchor and exceeds the 4 anchor, whose 'when' is less specific.

Trigger Term Quality

Natural user phrasings are present: 「表在哪」「哪个视图」「数据资产归属」「谁负责这类数据」「找表/找数」 — good keyword coverage a user would actually say. Not a 5 because common variants like 「数据在哪」「哪个库」「字段含义」 or asset-location synonyms are missing, leaving a few natural terms uncovered.

Distinctiveness Conflict Risk

The niche is clear (table/data-asset location in ontology knowledge networks, Chinese 找表/找数 questions) with triggers distinct from sibling data-retrieval skills, so it is mostly distinct. Minor overlap risk remains with closely related lookup skills (e.g., smart-ask-data for 指标/SQL 取数) whose questions could co-occur with 「数据资产归属」 phrasing — not a 5 because that boundary is not drawn in the description itself.