← Back to Skills Marketplace
Data Completeness Check
by
mirowangl-ops
· GitHub ↗
· v1.0.0
· MIT-0
103
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install xiaozhua-data-completeness-check
Description
数据完整性判断机制。通过前晚基准值 vs 早上值对比,判断 Power BI 数据是否回灌完毕。适用于任何有延迟回灌的数据源。避免发送不完整数据。
README (SKILL.md)
数据完整性判断
方法:双时间点对比
步骤
- 前一天晚上 23:30 — 抓取关键指标基准值(如 Paid Subs)
- 第二天早上 9:00 — 再抓一次同一指标
- 对比判断:
- 两次值不同 → 数据有更新,可能完整
- 两次值相同 → 数据没跑完,继续等
- 达到经验参考值 → 大概率完整
经验参考值(O+ 业务)
- Paid Subs 工作日 ~14,000
- Paid Subs 周末/节假日 ~10,000
- Revenue 工作日 ~$20,000-25,000
决策规则
- 达标 + 与昨晚不同 → ✅ 数据完整,直接发简报
- 达标但与昨晚相同 → ⚠️ 可能没跑完,等 30 分钟重试
- 未达标 → ❌ 先发 Lei 确认,不直接发群
通用化
此方法适用于任何有 T+1 延迟的数据源:
- 选一个敏感指标作为"哨兵"
- 建立该指标的经验参考范围
- 用双时间点对比判断是否回灌完毕
Usage Guidance
This skill is high-level and coherent, but check these practical points before enabling it: 1) Decide and document how metrics will be fetched (Power BI API, DB query, exported file) and ensure any credentials used are minimal-scope and provided separately from this skill. 2) If you allow the agent to call connectors autonomously, confirm which tokens/accounts it will use and limit their scope. 3) Define and record where the sampled values and logs will be stored and who can view them. 4) Test the procedure on non-sensitive data to validate thresholds and retry logic (30-minute wait) to avoid false positives/negatives. 5) If you need the skill to perform automated fetches, consider adding explicit instructions about the connector and required permissions so users understand what access will be needed.
Capability Analysis
Type: OpenClaw Skill
Name: xiaozhua-data-completeness-check
Version: 1.0.0
The skill bundle consists of logical instructions for an AI agent to perform data completeness checks by comparing metrics at different time intervals. It contains no executable code, network requests, or suspicious instructions, and its behavior is entirely aligned with the stated purpose of validating Power BI data integrity in SKILL.md.
Capability Assessment
Purpose & Capability
The name/description (data completeness check for Power BI / T+1 sources) matches the instructions (capture a sentinel metric at 23:30 and 09:00 and compare). Nothing requested (no env vars, no binaries, no installs) is disproportionate to that stated purpose.
Instruction Scope
The SKILL.md is high-level and prescribes a manual/automated sampling workflow (capture metric at two times, compare, use thresholds). It does not specify HOW to fetch metrics (e.g., Power BI API, DB query, or exported file) or where results are recorded. This ambiguity is not inherently malicious but means implementers must decide which connector/credentials to use.
Install Mechanism
No install spec and no code files are present (instruction-only), so nothing is written to disk and no third-party packages are pulled. This is the lowest-risk install surface.
Credentials
The skill declares no required environment variables, secrets, or config paths. The described task could require Power BI or database credentials in practice, but those are not requested by the skill itself—so the declared access is proportionate.
Persistence & Privilege
always is false and the skill does not request persistent presence or modify other skills or agent-wide settings. Autonomous invocation is allowed by platform default but is not excessive given the simple nature of the instructions.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install xiaozhua-data-completeness-check - After installation, invoke the skill by name or use
/xiaozhua-data-completeness-check - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
- Introduced a data completeness checking mechanism based on dual time-point comparison (late night vs. morning values).
- Outlined decision rules to verify if Power BI data has been fully updated before sending reports, preventing incomplete data delivery.
- Provided practical reference thresholds for key metrics (e.g., Paid Subs, Revenue) on weekdays and weekends/holidays.
- Made the approach generalizable for any delayed (T+1) data sources by selecting a key "sentinel" metric and establishing experience-based reference ranges.
Metadata
Frequently Asked Questions
What is Data Completeness Check?
数据完整性判断机制。通过前晚基准值 vs 早上值对比,判断 Power BI 数据是否回灌完毕。适用于任何有延迟回灌的数据源。避免发送不完整数据。 It is an AI Agent Skill for Claude Code / OpenClaw, with 103 downloads so far.
How do I install Data Completeness Check?
Run "/install xiaozhua-data-completeness-check" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Data Completeness Check free?
Yes, Data Completeness Check is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Data Completeness Check support?
Data Completeness Check is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Data Completeness Check?
It is built and maintained by mirowangl-ops (@mirowangl-ops); the current version is v1.0.0.
More Skills