pandas-pro

Performs pandas DataFrame operations for data analysis, manipulation, and transformation. Use when working with pandas DataFrames, data cleaning, aggregation, merging, or time series analysis. Invoke for data manipulation tasks such as joining DataFrames on multiple keys, pivoting tables, resampling time series, handling NaN values with interpolation or forward-fill, groupby aggregations, type conversion, or performance optimization of large datasets.

1.11x

Quality

100%

Does it follow best practices?

Impact

78%

1.11x

Average score across 6 eval scenarios

Securityby

Passed

No known issues

Evaluation results

86%

Customer Data Quality Remediation

Data cleaning pipeline

Criteria

Without context

With context

No silent dropna

100%

Empty string to NaN

100%

Uses .copy() on subset

50%

Method chaining

37%

100%

Nullable int type

100%

Categorical conversion

Vectorized string ops

100%

No iterrows

100%

Pre/post validation

100%

No chained indexing

100%

na_values on read

100%

Regional Sales Performance Summary

GroupBy aggregation with memory optimization

Criteria

Without context

With context

Categorical dtype

100%

observed=True

100%

Named aggregation

100%

Built-in aggregations

100%

No iterrows

100%

Memory usage check

100%

Vectorized column ops

100%

No chained indexing

100%

Single groupby call

100%

reset_index after groupby

100%

Filter before compute

100%

78%

23%

Order Reconciliation and Enrichment

Multi-DataFrame merging and validation

Criteria

Without context

With context

Merge validate param

100%

indicator=True usage

100%

Unmatched rows reported

50%

100%

pd.concat not append

100%

80%

No chained indexing

100%

Uses .loc for assignment

100%

75%

Downcast or categorical

reset_index after merge

50%

75%

No iterrows

50%

Meaningful suffixes

100%

Null check after join

100%

43%

Weather Station Data Pipeline

Time series resampling and gap filling

Criteria

Without context

With context

Resample via set_index

fillna(0) after resample

ffill then interpolate

25%

Mode fill for categoricals

Median fill for numerics

Pre-transform null check

100%

Post-transform null check

100%

No iterrows for time ops

100%

Dtype check at start

Aggregation uses built-in methods

62%

75%

No chained indexing

100%

73%

10%

Quarterly Sales Pivot Report

Pivot table reporting with assertion-based validation

Criteria

Without context

With context

pivot_table with fill_value

100%

margins=True

100%

Assert row count

Assert no nulls

100%

Assert column set

Initial data assessment

100%

Categorical dtype for pivot dims

No iterrows

100%

Vectorized derived columns

100%

No chained indexing

100%

Missing values handled explicitly

100%

90%

E-Commerce Transaction Log Processor

Chunked large dataset processing and memory optimization

Criteria

Without context

With context

Chunked CSV reading

100%

dtype spec on read

pd.concat to combine chunks

100%

Downcast integer columns

100%

Downcast float columns

37%

100%

Categorical for low-cardinality

100%

Memory usage reported

100%

Parquet output

100%

Filter before compute

100%

No iterrows

100%

No chained indexing

100%

Repository: jeffallan/claude-skills
Commit: 5b76101

Evaluated: 27 days ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Customer Data Quality Remediation Regional Sales Performance Summary Order Reconciliation and Enrichment Weather Station Data Pipeline Quarterly Sales Pivot Report E-Commerce Transaction Log Processor

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.