JPeetz/agent-skills / skills/data/data-analysis
data-analysis
VerifiedUse this skill when analyzing structured datasets: load, clean, explore, visualize, and report on data from CSV, JSON, Excel, or SQL sources. Produces statistical summaries, correlation matrices, time series analysis, regression models, hypothesis tests, and publication-quality visualizations.
| File | Kind | Size |
|---|---|---|
| CHANGELOG.md | markdown | 719 B |
| LICENSE | other | 1.0 KB |
| SKILL.md | skill md | 12.1 KB |
| evals/eval_cases.json | config | 6.2 KB |
| evals/evals.json | config | 3.3 KB |
| references/data-cleaning-guide.md | markdown | 8.3 KB |
| references/statistical-methods.md | markdown | 13.3 KB |
| references/visualization-patterns.md | markdown | 12.2 KB |
| scripts/validate_dataset.py | script | 20.5 KB |
evals/eval_cases.json
View on GitHub{
"skill": "data-analysis",
"version": "1.0.0",
"description": "Eval cases for the data-analysis agent skill. Positive cases should trigger the skill; near-miss negatives should not.",
"eval_cases": [
{
"id": "pos-001",
"type": "positive",
"label": "CSV analysis with charts",
"input": "I have a file called sales_2025.csv with columns date, product, region, revenue, units_sold. Can you analyze it and show me which products are performing best by region? Please include some charts.",
"expected_behavior": "Skill triggers. Agent loads the CSV, inspects structure, performs grouped analysis by product and region, creates bar charts showing revenue/units by region-product, and produces a structured report.",
"key_indicators": ["analyze", "CSV", "charts", "revenue", "by region"],
"required_phases": ["load", "inspect", "analyze", "visualize", "report"]
},
{
"id": "pos-002",
"type": "positive",
"label": "JSON data cleaning",
"input": "I scraped some product data and saved it as products.json. A lot of fields are missing or have weird values. Can you clean it up and tell me what's wrong with the data?",
"expected_behavior": "Skill triggers. Agent loads JSON, runs data quality checks (missing values, type inconsistencies, outliers), reports all issues, cleans the data with user confirmation for major decisions, and saves cleaned version.",
"key_indicators": ["clean", "JSON", "missing", "data quality", "weird values"],
"required_phases": ["load", "inspect", "clean", "report"]
},
{
"id": "pos-003",
"type": "positive",
"label": "Time series analysis",
"input": "Here's my website traffic data for the last 2 years in traffic.csv. I want to understand the trend, seasonality, and if there's been significant growth. Can you do a time series analysis?",
"expected_behavior": "Skill triggers. Agent loads CSV, parses date column, sets datetime index, performs time series decomposition (trend/seasonal/residual), calculates growth rates, creates line chart with trend overlay, and reports findings.",
"key_indicators": ["time series", "trend", "seasonality", "growth", "traffic"],
"required_phases": ["load", "inspect", "clean", "analyze", "visualize", "report"]
},
{
"id": "pos-004",
"type": "positive",
"label": "Correlation matrix and heatmap",
"input": "I have a dataset survey_results.csv with 20 numeric columns from a customer satisfaction survey. I want to see which factors correlate most strongly with overall satisfaction. Show me a correlation matrix.",
"expected_behavior": "Skill triggers. Agent loads CSV, computes correlation matrix, identifies strongest correlations with satisfaction score, creates a heatmap visualization, and reports top correlated factors with their correlation coefficients.",
"key_indicators": ["correlation", "heatmap", "numeric columns", "satisfaction", "factors"],
"required_phases": ["load", "inspect", "analyze", "visualize", "report"]
},
{
"id": "pos-005",
"type": "positive",
"label": "SQL query results analysis",
"input": "I ran a SQL query and exported the results to quarterly_revenue.csv. It has customer segments, quarters, and revenue numbers. Can you run a statistical test to see if there's a significant difference in revenue between enterprise and SMB customers?",
"expected_behavior": "Skill triggers. Agent loads CSV, splits data by customer segment (enterprise vs SMB), checks normality assumptions, runs appropriate statistical test (t-test or Mann-Whitney U), reports test statistic, p-value, effect size, and creates a box plot comparing the two groups.",
"key_indicators": ["SQL", "statistical test", "significant difference", "revenue", "CSV"],
"required_phases": ["load", "inspect", "analyze", "visualize", "report"]
},
{
"id": "neg-001",
"type": "near-miss-negative",
"label": "Database schema design (no data)",
"input": "I'm building a new e-commerce app. What columns should I include in my orders table? How should I structure the database schema for tracking inventory?",
"expected_behavior": "Skill does NOT trigger. This is a database schema design question without any actual data to analyze. Redirect to SQL/database design guidance instead.",
"key_indicators": ["schema design", "columns", "table", "no data file"],
"why_negative": "No dataset provided, no analysis to perform. This is schema architecture, not data analysis."
},
{
"id": "neg-002",
"type": "near-miss-negative",
"label": "General math theory question",
"input": "Can you explain how linear regression works? What's the difference between R-squared and adjusted R-squared? I'm studying for a statistics exam.",
"expected_behavior": "Skill does NOT trigger. This is a theoretical statistics question with no dataset context. Provide the explanation as general knowledge, not as a data analysis workflow.",
"key_indicators": ["explain how", "difference between", "studying", "exam", "no dataset"],
"why_negative": "Pure theory question. No data file, no analysis request, no visualization needed."
},
{
"id": "neg-003",
"type": "near-miss-negative",
"label": "Spreadsheet UI feature question",
"input": "How do I create a pivot table in Google Sheets? Also, I need to freeze the header row and apply conditional formatting to highlight cells above 100.",
"expected_behavior": "Skill does NOT trigger. This is a spreadsheet software UI/how-to question. Provide guidance on Google Sheets features without engaging the data analysis workflow.",
"key_indicators": ["Google Sheets", "pivot table", "freeze", "conditional formatting", "UI"],
"why_negative": "Spreadsheet UI operations, not programmatic data analysis. No dataset file provided for analysis."
}
],
"metrics": {
"total_cases": 8,
"positive_cases": 5,
"negative_cases": 3,
"evaluation_guidance": "For positive cases: verify that the skill triggers and follows at least the required phases. For negative cases: verify that the skill does NOT trigger and the response avoids launching the full analysis workflow."
}
}
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the work | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixes | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation code | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |