CtrlK
BlogDocsLog inGet started
Tessl Logo

data-science-python

Python data science: notebook structure, data validation, reproducibility, and model documentation

69

1.11x
Quality

60%

Does it follow best practices?

Impact

76%

1.11x

Average score across 3 eval scenarios

SecuritybySnyk

Passed

No known issues

Optimize this skill with Tessl

npx tessl skill review --optimize ./skills/data-science-python/SKILL.md
SKILL.md
Quality
Evals
Security

Evaluation results

57%

-6%

Predicting House Prices: Build a Jupyter Notebook Analysis

Notebook structure and reproducibility

Criteria
Without context
With context

Imports section first

100%

100%

Data loading second

100%

100%

EDA section present

100%

100%

Feature engineering section

100%

100%

Modeling section

100%

71%

Evaluation section

100%

71%

Conclusions section last

100%

100%

SEED constant defined

0%

0%

random.seed called with SEED

0%

0%

np.random.seed called with SEED

50%

50%

random_state=SEED on estimators

70%

60%

pathlib.Path for file access

0%

0%

82%

32%

Customer Churn Prediction Pipeline

Production pipeline with pandera and sklearn

Criteria
Without context
With context

Pandera import

0%

100%

Pandera DataFrameSchema

0%

100%

Schema validate called

0%

100%

sklearn Pipeline used

100%

100%

logging module used

100%

100%

No print in production code

100%

100%

Model filename has timestamp

0%

0%

Model filename has metadata

0%

0%

pathlib.Path for paths

0%

100%

Pinned requirements

100%

100%

pipeline.log produced

100%

100%

Model artifact exists

100%

100%

90%

-2%

Credit Risk Scoring: Model Card and Experiment Registry

Model documentation and experiment tracking

Criteria
Without context
With context

src/ module created

100%

100%

tests/ directory created

100%

100%

Unit tests for src functions

100%

100%

experiments/ directory exists

100%

100%

Three experiment JSON files

100%

71%

Experiment JSON has metadata

100%

100%

Model card: training data section

100%

100%

Model card: feature list

100%

100%

Model card: hyperparameters

100%

100%

Model card: evaluation metrics

100%

100%

Model card: limitations section

100%

100%

Pinned requirements

100%

100%

pathlib.Path used

38%

38%

Repository
ucdavis/ai-skills-registry
Evaluated
Agent
Claude Code
Model
Claude Sonnet 4.6

Table of Contents

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.