Token导航 LogoToken导航TokenDH.com
研究检索只读github未标认证来源可访问clear审计通过

altinity-expert-clickhouse-schemaAltinity 专家 CLIckhouse 架构

Agent Skill

altinity-expert-clickhouse-schema 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

1,572

周安装

63

GitHub Stars

5

下载量

509
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

3

许可证

MIT

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:altinity-expert-clickhouse-schema(Altinity 专家 CLIckhouse 架构)
来源仓库:https://github.com/altinity/skills
仓库路径:skills/altinity-expert-clickhouse-schema
安装命令:
npx skills add https://github.com/altinity/skills --skill altinity-expert-clickhouse-schema
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。不同来源提供的安装方式可能略有差异;本站展示可直接复制的安装命令,安装前请核对来源页面。

skills.shnpx skills
npx skills add https://github.com/altinity/skills --skill altinity-expert-clickhouse-schema

简介

评估 ClickHouse 表结构设计合理性并提出改进建议。

  • 适用于分析分区分布、ORDER BY 选择与物化视图影响。
  • 提供分区大小均衡度与 rows per partition 统计指标。
  • 依赖 checks.sql 获取表级元数据与性能指标基线。
  • 操作前应确认对 system.parts、system.columns 等表的查询权限。

SKILL.md

Table Schema and Design Analysis

Analyze table structure, partitioning, ORDER BY, materialized views, and identify design anti-patterns.


Diagnostics

Run all queries from the file checks.sql and analyze the results.


Deep Dive Queries (Placeholder-Based)

Partition Distribution for Specific Table

select
    database,
    table,
    count() as partitions,
    sum(rows) as total_rows,
    formatReadableSize(sum(bytes_on_disk)) as total_size,
    formatReadableSize(median(bytes_on_disk)) as median_partition_size,
    min(partition) as oldest_partition,
    max(partition) as newest_partition
from system.parts
where active and database = '{database}' and table = '{table}'
group by database, table, partition
order by partition desc
limit 100

Column Compression Analysis for Specific Table

select
    name,
    type,
    formatReadableSize(data_compressed_bytes) as compressed,
    formatReadableSize(data_uncompressed_bytes) as uncompressed,
    round(data_uncompressed_bytes / nullIf(data_compressed_bytes, 0), 2) as ratio,
    compression_codec
from system.columns
where database = '{database}' and table = '{table}'
order by data_compressed_bytes desc
limit 50

Look for:

  • Columns with ratio < 2 → consider better codec or data transformation
  • Large columns without codec → add CODEC(ZSTD) or LZ4HC
  • String columns with low cardinality → consider LowCardinality(String)

Index Usage Analysis for Specific Database

select
    database,
    table,
    name as index_name,
    type,
    expr,
    granularity
from system.data_skipping_indices
where database = '{database}'
order by database, table

Schema Design Recommendations

Partition Key Guidelines

Data VolumeRecommended GranularityExample
< 10GB/monthNo partitioning or yearlytoYear(ts)
10-100GB/monthMonthlytoYYYYMM(ts)
100GB-1TB/monthWeekly or dailytoMonday(ts)
> 1TB/monthDailytoDate(ts)

ORDER BY Guidelines

  1. First column: Low cardinality, frequently filtered (e.g., tenant_id, region)
  2. Second column: Time-based if range queries common
  3. Subsequent: Other filter columns by selectivity (most selective last)

Anti-patterns:

  • UUID/hash as first column
  • High-cardinality ID without tenant prefix
  • DateTime64 with microseconds as first column

Compression Codec Recommendations

Data TypeRecommended Codec
Integers (sequential)Delta, ZSTD
Integers (random)ZSTD or LZ4HC
FloatsGorilla, ZSTD
TimestampsDoubleDelta, ZSTD
Strings (long)ZSTD(3)
Strings (repetitive)LowCardinality + ZSTD

Cross-Module Triggers

FindingLoad ModuleReason
Many small partitionsaltinity-expert-clickhouse-ingestionCheck batch sizing
Oversized partitionsaltinity-expert-clickhouse-mergesMerge can't complete
High PK memoryaltinity-expert-clickhouse-memoryMemory pressure
MV performance issuesaltinity-expert-clickhouse-reportingQuery analysis
Too many parts per partitionaltinity-expert-clickhouse-mergesMerge backlog

Settings Reference

SettingDefaultRecommendation
index_granularity8192Lower for point lookups, higher for scans
ttl_only_drop_parts0Set to 1 if TTL deletes entire partitions
min_bytes_for_wide_part10MBIncrease if many small parts

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

04

需要参考平台分布和安装热度时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

补充不同宿主或平台的使用分布数据

能力 5

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Codex

27.66%
按下载量换算141

Claude Code

23.16%
按下载量换算118

Gemini CLI

17.81%
按下载量换算91

Antigravity

13.99%
按下载量换算71

cline

9.14%
按下载量换算47

OpenCode

3.47%
按下载量换算18

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。

来源信息

继续浏览同类 Skills