data-analyst

What This Skill Does

The data-analyst skill transforms your AI agent into a sophisticated data processing engine capable of bridging the gap between raw datasets and actionable business intelligence. It excels at parsing structured and semi-structured data from various origins, including SQL relational databases, CSV files, Excel spreadsheets, and Google Sheets. Beyond simple extraction, the skill empowers the agent to perform complex statistical analysis, including descriptive statistics, trend correlation, and outlier detection. It further supports automated report generation and high-quality data visualization, allowing you to turn complex query results into intuitive charts and dashboard components. Whether you are performing cohort analysis, conversion funnels, or routine data cleaning, this skill acts as your automated data scientist.

Installation

To integrate this skill into your environment, use the command-line interface to pull from the OpenClaw registry:

clawhub install openclaw/skills/skills/oyi77/data-analyst

After installation, ensure your environment variables and connection strings are correctly mapped in your TOOLS.md file. Specifically, verify that you have provided the necessary credentials for your databases or cloud storage providers. Once configured, initialize the workspace by running ./scripts/data-init.sh to ensure all dependencies are resolved.

Use Cases

1. Automated Business Intelligence: Scheduling recurring reports that pull performance metrics from a database and summarize key trends for management.
2. Database Health Monitoring: Identifying missing values, duplicate entries, or schema drift within large production datasets.
3. Product Analytics: Calculating conversion funnels, user retention rates, and churn statistics without writing manual dashboard code.
4. Spreadsheet Automation: Normalizing messy, user-submitted Excel files before merging them into a central system of record.

Example Prompts

1. "Look at the last 30 days of transactions in the postgres database, group them by product category, and generate a bar chart showing which categories had the highest revenue growth month-over-month."
2. "Analyze the 'customer_feedback.csv' file in my local folder. Clean up the date formatting, remove any entries with empty sentiment scores, and provide a summary of the most frequent keywords found in the comments."
3. "Write a SQL query to identify our cohort of users who signed up in January 2023 and track their purchase activity over the subsequent six months. Explain the churn rate for this specific group."

Tips & Limitations

• Security First: Always use read-only database roles when connecting your agent to production databases to prevent accidental data modification.
• Data Privacy: Be mindful of PII (Personally Identifiable Information) when uploading spreadsheets. Use the data cleaning functions to anonymize sensitive columns.
• Performance: For extremely large datasets (multi-gigabyte), ensure your database has proper indexing before running complex JOINs or time-series aggregations.
• Refinement: If the agent's SQL output is slow, ask it to add limits or filters early in the query process to optimize execution time.