Data Quality Frameworks
Use when implementing data-quality checks, data contracts, validation pipelines, freshness monitoring, or CI/CD gates for analytical and operational datasets.
Install
npx skills add sandbaseai/workbuddy-skill --skill data-quality-frameworksgit clone --depth 1 https://github.com/sandbaseai/workbuddy-skill.git /tmp/workbuddy-skill
mkdir -p ~/.claude/skills
cp -r /tmp/workbuddy-skill/skills/data-quality-frameworks ~/.claude/skills/data-quality-frameworksPersonal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.
From the SKILL.md
# Data Quality Frameworks 为数据管道建立可解释、可重复、隐私安全的质量验证。围绕数据契约和业务关键字段检查完整性、唯一性、有效性、准确性、一致性和时效性,并将失败转化为有证据的阻断、告警或人工复核决定。 ## 范围和安全边界 - 开始前确认数据集、版本/分区、来源、时区、schema、业务负责人、允许的样本范围和质量目标。 - 默认只读检查 schema、元数据、统计摘要和授权测试数据;不写入生产表、不删除/隔离记录、不触发真实告警或重跑高成本全量任务。 - 采样必须有界并记录方法、覆盖率、偏差和查询成本;抽样通过不能推断全量通过。 - 记录聚合值、计数、比例和脱敏示例,不保存 PII、支付信息、秘密、完整原始行或可还原的用户标识。 - 质量失败不自动等同于业务事故;区分数据问题、契约变更、上游延迟、采集缺失、规则过时和验证器故障。 ## 数据契约 为每个关键数据集建立版本化契约,至少包含:字段名/类型/可空性、允许值或范围、主键/唯一性、关系约束、分区/时区、新鲜度目标、敏感级别、所有者、兼容性策略和失败处置。契约变更必须说明新增、删除、收窄、放宽和迁移窗口;没有批准版本时标记 `unknown`,不从当前样本反推契约。 ## 质量维度与检查 按风险选择检查,不盲目测试每一列: | 维度 | 典型问题 | 证据 | |---|---|---| | 完整性 | 关键字段空值、分区缺失 | null ratio、分区计数 | | 唯一性 | 主键或业务键重复 | distinct/duplicate count | | 有效性 | 枚举、格式、范围非法 | 合法值比例、越界摘要 | | 准确性 | 与可信源不一致 | 对账差异、抽样交叉核对 | | 一致性 | 跨字段/跨表矛盾 | 关系失败计数、约束摘要 | | 时效性 | 数据延迟或陈旧 | 最大事件时间、延迟分布 | 每条规则写清字段、谓词、单位、适用分区、空集行为、允许误差、严重性和失败动作。`null`、未知枚举、重复键、无数据和验证器错误不能混为一个 0 分。 ## 基线、阈值和增量验证 优先在源数据进入转换前检查,再逐层检查关键转换和跨表关系。阈值以历史基线、业务 SLA 或契约为依据,记录窗口、样本量、统计方法和置信限制;不要硬编码没有来源的阈值。增量检查覆盖新增/变更分区,并定期抽查历史分区,避免旧数据永久逃逸。动态基线必须设置上下限、冷启动、季节性和异常缺失策略。 检查计算失败率和“未执行”状态,避免把空输入或查询超时当成全通过。高风险字段、跨系统对账、权限/租户隔离和金额等规则应设置阻断级别;低风险漂移可以告警但仍需记录。 ## 流水线和报告 在 CI/CD 或编排中把质量检查作为有界步骤:固定数据版本/分区,输出机器可读结果和人类摘要,保留规则版本、运行时间、查询 ID、样本统计和脱敏证据。失败报告至少包含 dataset、partition、rule、expected、observed、severity、coverage、baseline、owner、下一动作和是否阻断发布。 失败处置分为:阻断下游、重试上游读取、延迟等待 freshness、进入人工复核、发布兼容变更或登记已批准例外。重试要有次数和总时限;例外必须有负责人、理由、范围、到期日和补偿检查,不得永久吞掉失败。 ## 监控与演进
Details
- Skill name
data-quality-frameworks- Repository
- sandbaseai/workbuddy-skill
- Path
skills/data-quality-frameworks/SKILL.md- License
- MIT
- Repo updated
- 8 September 2026
FAQ
What does the Data Quality Frameworks skill do?
Use when implementing data-quality checks, data contracts, validation pipelines, freshness monitoring, or CI/CD gates for analytical and operational datasets.
How do I install the Data Quality Frameworks skill?
skills CLI (Claude Code, Cursor, Codex and more): npx skills add sandbaseai/workbuddy-skill --skill data-quality-frameworks. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.
Is Data Quality Frameworks popular?
It lives in sandbaseai/workbuddy-skill, which has 9 GitHub stars and 3 forks.
Related skills
Access Control Security Best Practices
9sandbaseai/workbuddy-skill
Design, review, implement, and verify least-privilege access controls across users, services, tenants, resources, and administrative workflows.
Accessibility Review
9sandbaseai/workbuddy-skill
Use when reviewing a web interface or interaction for semantic, keyboard, focus, screen-reader, ARIA, and contrast accessibility risks.
Acquire Codebase Knowledge
9sandbaseai/workbuddy-skill
Use when the user asks to map, document, or onboard into an existing codebase and needs evidence-backed architecture, structure, conventions, integrations, testing, and concerns documentation.
Debug Triage
5Sagargupta16/claude-skills
Use when triaging a pasted error, stack trace, or terminal transcript. Activates on "debug this", "fix this", "why is this broken", or when a user pastes red terminal output. Produces a tight cause-fix-test report, not a lecture.
Constraint Driven Development
9sandbaseai/workbuddy-skill
Use when a project needs a written, measurable quality bar that agents and CI can enforce without quietly weakening tests, security, performance, accessibility, or architecture checks.
Debugging Methodology
9sandbaseai/workbuddy-skill
Use when investigating a software defect, intermittent failure, incorrect result, crash, or unexplained operational symptom.
Browse the most popular Claude skills, all software development skills, or write your own with the skills generator.