Data Science Agent Skills for Analysis
Data science skills cover the full arc of applied analysis with models in the mix: framing the question, deciding what data you may train and test on, shaping the representation the model actually sees, choosing a model and a dumb baseline it has to beat, evaluating with numbers that would still hold next month, and delivering a prediction or insight someone else can use. Skills in this topic tend to be narrow — one that helps you interrogate whether a better prediction would change a decision, another that guards your train/test split, another that builds a baseline before anything fancier. Use the ranked list by starting at the top and reading each skill's description against your current bottleneck. If you are still arguing about the question, skip the modeling entries. If you have a model and no baseline, start there. Pick the one skill that unblocks the stage you are actually in.
8 to start with16 top options
Start with these.
Data science is applied analysis with models in the mix: question, data, features, model, evaluate, then put it somewhere it can be used.
These are the closest matches for this work. Start with the first one and open it to install.
- Rank 01View Agent SkillDirect
statistical-analysis
Applies statistical methods including descriptive statistics, trend analysis, outlier detection, hypothesis testing, and forecasting, with guidance on cautious interpretation of statistical claims.
4,314 · popularity - Rank 02View Agent SkillDirect
ai-research-explore
Runs candidate-only deep learning research exploration on top of a current_research anchor, with idea gating, fair comparison, and governed experiments written to explore_outputs.
311,409 · popularity - Rank 03View Agent SkillDirect
datanalysis-credit-risk
Credit risk data cleaning and variable screening pipeline for pre-loan modeling, covering missing rates, IV/PSI filtering, null-importance denoising, and correlation removal.
7,300 · popularity - Rank 04View Agent SkillDirect
data-analysis
Analyzes uploaded Excel/CSV files using DuckDB, supporting schema inspection, SQL querying, statistical summaries, multi-sheet workbooks, and result export.
3,544 · popularity - Rank 05View Agent SkillDirect
nature-statistics
Audits and improves manuscript statistical reporting, covering experimental units, replication, uncertainty, tests, and figure statistics, with conservative ready-to-paste text.
10,201 · popularity - Rank 06View Agent SkillDirect
dbt-transformation-patterns
Covers dbt analytics engineering with model organization into staging/intermediate/marts layers, data quality testing, documentation, and incremental processing strategies.
10,521 · popularity - Rank 07View Agent SkillDirect
azure-kusto
OfficialQueries and analyzes data in Azure Data Explorer (Kusto/ADX) using KQL for log analytics, telemetry, time series analysis, aggregations, and anomaly detection.
544,362 · popularity - Rank 08View Agent SkillDirect
powerbi-modeling
Guides Power BI semantic modeling with star schemas, DAX measures, relationships, RLS, naming conventions, and model performance optimization.
Needs (stated): MCP server
9,868 · popularity
Explore 8 more Agent Skills
- Rank 09View Agent SkillDirect
data-storytelling
Transforms data into narratives using visualization, context, and persuasive structure for presenting analytics to stakeholders and building executive data reports.
15,981 · popularity - Rank 10View Agent SkillDirect
clickhouse-io
Covers ClickHouse schema design, query optimization, aggregations, window functions, and data engineering patterns for high-performance analytical workloads.
9,781 · popularity - Rank 11View Agent SkillPossible
ai-research-reproduction
Guides README-first deep learning repository reproduction toward a minimal trustworthy run, coordinating intake, setup, execution, training verification, and evidence recording.
311,091 · popularity - Rank 12View Agent SkillPossible
content-experimentation-best-practices
Covers content experimentation and A/B testing with experiment design, hypotheses, metrics, sample size, statistical foundations, and common analysis pitfalls.
3,615 · popularity - Rank 13View Agent SkillPossible
ab-testing
Plans and designs A/B and multivariate experiments with hypothesis frameworks, sample size tables, metric selection, and statistical significance guidance.
68,676 · popularity - Rank 14View Agent SkillDirect
bigquery-ai-ml
Official>- Leverages BigQuery's built-in machine learning and GenAI capabilities for advanced data analytics. Use when you need to write SQL queries that perform time-series forecasting, predict values, detect outliers or anomalies, find key drivers, perform semantic search or vector search, classify text, calculate similarity, summarize content, translate language, evaluate models, filter by semantic conditions, or leverage generative AI capabilities in BigQuery. Do not use for general BigQuery dataset, table, or job management requests.
- Rank 15View Agent SkillDirect
finetuning
OfficialFine-tune models on Microsoft Foundry using SFT (supervised), DPO (preference), or RFT (reinforcement with graders). Covers dataset preparation, training job submission, deployment, and evaluation. USE FOR: fine-tune, SFT, DPO, RFT, training data, grader, distillation, fine-tuned model, training job, large file upload, calibrate grader, deploy fine-tuned model, evaluate fine-tuned model. DO NOT USE FOR: general model deployment without fine-tuning (use deploy-model), agent creation (use agents), prompt optimization without training (use prompt-optimizer).
- Rank 16View Agent SkillDirect
google-cloud-solution-agentic-ai-data-science-workflow
Official>- Designs a tailored multi-product agentic data science architecture on Google Cloud that incorporates opinionated best practices. Use when architecting multi-product solutions for agent-based data analytics or ML workloads. Don't use for simple queries, non-agentic pipelines, general cloud reviews, or writing agent code.
Needs (stated): Network access
What this topic covers
Each sub-topic groups Agent Skills for a specific situation.
- Question
- What decision a better prediction would change.
- Data
- What you are allowed to train and test on.
- Features
- The representation the model actually sees.
- Model
- A model, and a dumb baseline it has to beat.
- Evaluate
- Numbers that would still hold next month.
- Deliver
- A prediction or an insight someone else can use.
Frequently asked questions
- How is the list ordered?
- It is ranked by observed search demand, with direct matches for this topic placed first and a cap on how many skills any single publisher can occupy.
- Do I need all six stages covered before a data science skill is useful?
- No. The stages are question, data, features, model, evaluate, and deliver, and most skills address one of them. A skill aimed at evaluation is useful even if your question framing is still rough.
- Where should I start if I already have a model?
- Check whether you have a dumb baseline the model has to beat, and whether your evaluation numbers would still hold next month. Skills in this topic that target those two stages are the natural next pick.
- What separates a data science skill from a general coding skill?
- Data science here means applied analysis with models in the mix, so the work runs through question, data, features, model, evaluate, and deliver. A skill that does not touch that arc belongs to a different topic.
How to install an Agent Skill
Start with the first skill on this page. Claude Code reads a SKILL.md once it is in the skills directory:
- Claude Code — copy the SKILL.md for statistical-analysis into ~/.claude/skills/statistical-analysis/ for yourself, or .claude/skills/statistical-analysis/ to share it with a project.
- From the registry — run npx skills add anthropics/knowledge-work-plugins@statistical-analysis.
- Or download statistical-analysis from its source and place that SKILL.md in the directory above.
How to choose Check the source and the skill’s own instructions before you install. Official and Notable describe where a skill comes from.