dashboard-review

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

dashboard-review

dashboard-review

When to use

使用场景

Use this skill when tasks include:
  • reviewing a PR that modifies Kibana dashboard JSON files
  • comparing dashboard changes between branches
  • summarizing what changed in dashboard exports
  • checking dashboard changes against the official dashboard guidelines
当你需要完成以下任务时使用本技能:
  • 审查修改了Kibana仪表盘JSON文件的PR
  • 对比不同分支间的仪表盘变更
  • 总结仪表盘导出文件中的变更内容
  • 检查仪表盘变更是否符合官方仪表盘规范

When not to use

非适用场景

Do not use this skill as the primary guide for:
  • creating new dashboards or exporting them from Kibana (
    dashboard-guidelines
    )
  • dashboard naming conventions and file layout (
    dashboard-guidelines
    references/kibana-assets-layout.md
    )
  • package-wide command orchestration (
    elastic-package-cli
    )
  • test suite selection (
    integration-testing
    )
请勿将本技能作为以下任务的主要指南:
  • 创建新仪表盘或从Kibana导出仪表盘(请使用
    dashboard-guidelines
  • 仪表盘命名规范与文件布局(请查看
    dashboard-guidelines
    references/kibana-assets-layout.md
  • 包级命令编排(请使用
    elastic-package-cli
  • 测试套件选择(请使用
    integration-testing

Prerequisites

前置条件

kbdash
must be installed:
bash
go install github.com/efd6/kbdash@latest
Lines prefixed with
[!]
in the output are warnings about potential issues in the dashboard JSON (consistency mismatches, incomplete extraction, parse errors). Run
kbdash -h
for a description of each warning type.
必须安装
kbdash
bash
go install github.com/efd6/kbdash@latest
输出中以
[!]
开头的行是关于仪表盘JSON中潜在问题的警告(如一致性不匹配、提取不完整、解析错误)。运行
kbdash -h
可查看每种警告类型的说明。

Review procedure

审查流程

Read
references/review-procedure.md
for the full step-by-step workflow. The high-level flow is:
  1. Identify changed dashboard files (added, removed, modified).
  2. Extract before/after descriptions with
    kbdash
    .
  3. Compare descriptions and classify changes as meaningful or cosmetic.
  4. Verify suspected issues against the raw JSON before reporting.
  5. Format output as one section per dashboard with bullet-pointed changes.
  6. Check guideline compliance on added or modified dashboards.
阅读
references/review-procedure.md
获取完整的分步工作流程。大致流程如下:
  1. 识别变更的仪表盘文件(新增、删除、修改)。
  2. 提取使用
    kbdash
    获取变更前后的描述。
  3. 对比描述内容,将变更分类为实质性变更或外观性变更。
  4. 验证在报告前,对照原始JSON核实疑似问题。
  5. 格式化输出格式为每个仪表盘对应一个章节,用项目符号列出变更内容。
  6. 检查新增或修改的仪表盘是否符合规范。

Guideline compliance checks

规范合规性检查

After summarizing changes, check the final state of every added or modified dashboard against the official dashboard guidelines. Report violations in a "Guideline notes" subsection after the change summary for each dashboard.
Check for these issues:
  • TSVB panels: Flag any
    visualization
    panel using TSVB. The guidelines require Lens for all new visualizations. Migrating existing TSVB to Lens is encouraged.
  • Missing dataset filter: Each visualization should filter on
    data_stream.dataset
    or an equivalently specific scope. Flag panels that query broad index patterns (
    metrics-*
    ,
    logs-*
    ) without scoping.
  • By-reference panels: Panels should be embedded by value. In the raw JSON,
    references
    entries with a
    panelRefName
    indicate by-reference panels. Flag these.
  • Deprecated input controls: The
    input-control-vis
    type is deprecated. Dashboard-native controls should be used instead.
  • Package-name title prefix: Panel titles matching
    [<Package Name> ...]
    create unnecessary repetition. Flag these.
  • Broad wildcard filters: Filters using unscoped
    -*
    patterns without further qualification are a performance concern.
  • High panel count: If a dashboard has more than roughly 20 panels, note it. The guidelines recommend splitting across dashboards and linking with drilldowns.
Only report issues that actually exist — skip passing checks. For pre-existing violations in unchanged panels, mention them once briefly ("N existing panels also use TSVB") rather than listing each one. Focus review attention on newly added or modified panels.
总结变更内容后,检查所有新增或修改的仪表盘的最终状态是否符合官方仪表盘规范。在每个仪表盘的变更总结后添加“规范说明”小节,报告违规情况。
需检查的问题:
  • TSVB面板: 标记所有使用TSVB的
    visualization
    面板。规范要求所有新可视化组件使用Lens。建议将现有TSVB迁移至Lens。
  • 缺失数据集过滤器: 每个可视化组件应基于
    data_stream.dataset
    或同等具体的范围进行过滤。标记未限定范围、直接查询宽泛索引模式(如
    metrics-*
    logs-*
    )的面板。
  • 引用式面板: 面板应通过值嵌入。在原始JSON中,带有
    panelRefName
    references
    条目表示引用式面板,需标记此类面板。
  • 已弃用的输入控件:
    input-control-vis
    类型已被弃用,应改用仪表盘原生控件。
  • 包名称标题前缀: 面板标题匹配
    [<Package Name> ...]
    格式会造成不必要的重复,需标记此类标题。
  • 宽泛通配符过滤器: 使用未限定范围的
    -*
    模式且无进一步限定条件的过滤器会影响性能,需标记。
  • 面板数量过多: 如果仪表盘包含约20个以上的面板,需记录。规范建议将其拆分到多个仪表盘,并通过钻取功能链接。
仅报告实际存在的问题——无需提及符合要求的检查项。对于未变更面板中已存在的违规情况,只需简要提及一次(如“已有N个面板使用TSVB”),无需逐个列出。审查重点应放在新增或修改的面板上。

Handoff to other skills

转交至其他技能

After the review is complete:
  1. dashboard-guidelines
    for creation or export guidance if the review surfaces structural issues
  2. package-spec
    when dashboard changes require a release note entry
  3. elastic-package-cli
    for validation commands
审查完成后:
  1. 若审查发现结构问题,可使用
    dashboard-guidelines
    获取创建或导出指导
  2. 当仪表盘变更需要发布说明条目时,使用
    package-spec
  3. 使用
    elastic-package-cli
    执行验证命令

References

参考资料

  • references/review-procedure.md
    — full extraction, comparison, and output formatting workflow
  • references/review-procedure.md
    — 完整的提取、对比和输出格式化工作流程