smart-ask-data

问数端到端编排（native CLI 版）：从候选 KN 选定知识网络，用 bkn object-type 发现对象类与字段，由编排层 LLM 生成 SQL，再由 ontology dataview query 执行取数；最后输出中文结论与口径说明。当用户需要指标、统计、趋势、SQL 取数或数据查询时使用。

Quality

80%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./configs/microservice/bff-service/configs/agent-skills/ontology/smart-ask-data/SKILL.md

Quality

Content

71%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is a well-structured orchestration spec: an ordered step checklist, concrete command templates, strict result-integrity rules, and cleanly split one-level-deep references. Its weaknesses are redundant restatement of the delegation/gateway/token contract across sections and the absence of an explicit error-recovery feedback loop.

Suggestions

State the ontology-core delegation, gateway/base-url, and no-token contract once in a single 委托契约 section and reference it elsewhere instead of restating it in 调用方式, 子技能依赖, 与 smart-data-analysis 的关系, and 配置 — this trims the conciseness duplication.

Add an explicit error-recovery step for failed dataview queries (e.g., on schema/SQL error: re-run schema discovery, regenerate SQL, retry once) to turn the implicit schema-first validation into a closed validate→fix→retry loop and lift workflow_clarity.

Provide at least one fully-resolved example command (with a concrete kn-id / dataview-id / accountId value) in tool-examples.md so the command templates read as copy-paste ready rather than placeholder-only.

Dimension	Reasoning	Score
Conciseness	The body is table- and checklist-driven with no over-explanation of known concepts, but the ontology-core delegation contract, the gateway/base-url note, and the no-token note are each restated across 4–5 sections (调用方式, 子技能依赖, 与 smart-data-analysis 的关系, 配置), which is unnecessary duplication that could be tightened — matching the 3 'mostly efficient but could be tightened' anchor rather than the 4 'minor instances'.	3 / 5
Actionability	Concrete command templates are given — 'bkn get <kn-id>', 'bkn object-type list/get', 'dataview query <dataview-id> --sql "…"' — with key options and ordering, but they use placeholders and the real execution entry is intentionally omitted ('不出现真实执行入口'), so they are mostly-but-not-fully executable, fitting the 4 anchor.	4 / 5
Workflow Clarity	A clearly ordered 5-step 主流程 checklist plus per-step 约束 (schema-first to prevent SQL hallucination, SELECT/WITH-only enforcement, empty-result handling) provide most checkpoints, but no explicit validate→fix→retry feedback loop for failed queries keeps it at 4 rather than 5.	4 / 5
Progressive Disclosure	The body is a clear overview that points to a '必读 references（按步骤）' table linking four one-level-deep reference files (kn-resolve.md, schema-discovery.md, sql-execute.md, tool-examples.md), all of which exist in ./references/, with links clearly signaled and organized by step — matching the 5 anchor.	5 / 5
	Total	16 / 20 Passed

Description

88%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is strong: it concretely lays out the full data-query orchestration pipeline with specific commands and provides an explicit natural-language trigger clause covering both what and when. Its main weakness is slightly broad trigger terms that overlap with sibling skills and a few missing common synonyms.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions tied to specific commands — '用 bkn object-type 发现对象类与字段', '由编排层 LLM 生成 SQL', 'ontology dataview query 执行取数', '输出中文结论与口径说明' — covering the full select→discover→generate→execute→output pipeline, matching the comprehensive-coverage anchor rather than the 4 anchor's 'minor gaps'.	5 / 5
Completeness	Both 'what' (the end-to-end orchestration pipeline) and 'when' (the explicit '当用户需要…时使用' trigger clause) are clearly and concretely answered, matching the 5 anchor; not the 4 anchor since the trigger is explicit rather than merely 'could be more specific'.	5 / 5
Trigger Term Quality	Explicit natural triggers appear in '当用户需要指标、统计、趋势、SQL 取数或数据查询时使用', giving good keyword coverage, but common domain synonyms like 报表/汇总/看数 are absent, so it stops short of the 5 anchor's comprehensive synonym coverage.	4 / 5
Distinctiveness Conflict Risk	The 'what' is a distinctive ontology-native-CLI 问数 orchestration niche, but the trigger terms (统计/趋势/数据查询) are broad enough to overlap with the sibling smart-data-analysis skill, fitting the 4 'mostly distinct; minor overlap risk with closely related skills' anchor rather than the 5 'minimal conflict risk'.	4 / 5
	Total	18 / 20 Passed

Validation

87%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
relative_links	Relative link issues: 1 missing, 6 suspicious	Warning

	Total	14 / 16 Passed

Repository: UnicomAI/wanwu
Path: configs/microservice/bff-service/configs/agent-skills/ontology/smart-ask-data/SKILL.md
Commit: 030c20b

Reviewed: 1 day ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.