📦 Data — 数据

v1.0.0

分析CSV文件的全面数据分析工作流,提供交互式引导与灵活输出格式。当用户提及“analyzin...”时即可调用此技能。

0· 18·0 当前·0 累计
下载技能包
最后更新
2026/4/21
0
安全扫描
VirusTotal
无害
查看报告
OpenClaw
安全
medium confidence
The skill's instructions and requirements are coherent for a CSV data‑analysis helper, but there are some minor gaps (undeclared Python dependencies and no install guidance) and user-safety considerations when giving arbitrary file paths.
评估建议
This skill appears to do what it says — analyze CSVs with pandas and produce reports — but before installing or invoking it: (1) Ask the skill author or installer which Python packages (pandas, numpy, plotting libraries, etc.) are required and whether the environment will install them automatically. (2) Be cautious when supplying file paths: the skill will read whatever path you provide, so do not point it at system or secret files (for example ~/.ssh, credential stores, or other non-CSV sensiti...
详细分析 ▾
用途与能力
The skill is described as a CSV data analysis workflow and the SKILL.md contains stepwise instructions for loading, cleaning, EDA, modeling and producing reports — all consistent with that purpose. Minor inconsistency: the instructions show Python code that uses pandas and numpy, but the skill declares no required binaries or dependencies and provides no install spec.
指令范围
Instructions explicitly tell the agent to load user-provided CSV paths and execute pandas-based analysis, include interactive decision points and an --auto mode. That stays within data-analysis scope. However, there are no safeguards against being asked to read non-CSV or sensitive system files (the agent will follow the path the user supplies). The evals reference test data under ~/.claude/skills/..., indicating expected file access within the agent's home directory; this is normal for data analysis tests but worth noting as it implies the skill will read arbitrary filesystem paths if provided.
安装机制
This is an instruction-only skill (no install spec), which is low risk. But the runtime snippets require Python libraries (pandas, numpy, Charting/tooling for HTML dashboards) that are not declared. Without those present in the runtime, the skill will fail or the agent may need to install them ad hoc — the skill gives no guidance on that.
凭证需求
The skill does not request environment variables, credentials, or config paths. That is proportionate to its stated purpose (local CSV analysis).
持久化与权限
always:false and no install persistence are appropriate. The skill does not request elevated or persistent privileges. Note: the agent can invoke the skill autonomously by default (platform default), which is expected for skills; this is not a new privilege here.
安全有层次,运行前请审查代码。

运行时依赖

无特殊依赖

版本

latestv1.0.02026/4/21

init

无害

安装命令

点击复制
官方npx clawhub@latest install yz6214589-hash-data-analysis
镜像加速npx clawhub@latest install yz6214589-hash-data-analysis --registry https://cn.longxiaskill.com
数据来源ClawHub ↗ · 中文优化:龙虾技能库