Data Analyst

**Quick judgment:** This is a strong, general-purpose data analysis skill that gives an OpenClaw agent most of what a human data analyst routinely does: SQL querying, spreadsheet analysis, visualizations, statistics, and structured reporting. It is well-suited as a default analytics capability for many business-focused agents. **Key strengths** - Covers the full pipeline from raw data (DBs, CSV/Excel/Sheets) to cleaned datasets, charts, and narrative reports. - Provides concrete SQL and pandas recipes for common business questions (trends, funnels, cohorts, correlations, data quality checks). - Good developer ergonomics: clear configuration via TOOLS.md and helper scripts for queries, analysis, and report generation. - Strong emphasis on best practices and methodology (analysis workflow, report templates, common mistakes to avoid). **Main risks / limitations** - **Data correctness risk:** The skill offers patterns, but the agent still must adapt SQL/pandas code to the actual schema and business logic; misalignment can produce subtly wrong analyses. - **No built-in access control:** Safety around PII/production data depends entirely on how data sources and permissions are configured externally. - **Scope bounded to classical analytics:** It does not natively cover heavy ML/advanced modeling or real-time/streaming analytics; those would require complementary skills. **Recommended scenarios** - Business, product, marketing, or operations agents that need repeatable, explainable analytics and reporting. - Internal tools that turn user-uploaded CSV/Excel files into dashboards and executive summaries. - Agents supporting data quality audits or migration validation for databases and warehouses. - As a foundational analytics layer to combine with more specialized skills (e.g., forecasting, experimentation, or domain-specific tooling).