Data and ML collection

Data Science and Machine Learning Skills for Claude Code and Antigravity

For analysts, data scientists and ML engineers who want their coding agent to analyze data, tune queries and train or run models. To install one, download it and put the skill folder (the one with SKILL.md) in your agent’s skills directory: .claude/skills/ for Claude Code or .agents/skills/ for Antigravity. The install guide covers global folders and CLI options.

Updated November 2026

At a glance

SkillBest forAuthor
Data AnalystBusiness data analysisAntigravity Skills
SQL OptimizationSlow SQL queriesAntigravity Skills
Excel MasterSpreadsheets and financial modelsAntigravity Skills
A/B Test SetupValid A/B experimentsAntigravity Skills
NetworkXGraph and network analysisAntigravity Skills
SymPyExact symbolic mathAntigravity Skills
Hugging Face CLIManaging Hub models and datasetsAntigravity Skills
Hugging Face Dataset ViewerExploring Hub datasetsAntigravity Skills
Hugging Face Model TrainerFine-tuning language modelsAntigravity Skills
Hugging Face JobsCloud compute without setupAntigravity Skills
Hugging Face TrackioExperiment trackingAntigravity Skills
Transformers.jsML inference in JavaScriptAntigravity Skills

Which data skill fits your job?

The 12 data and ML skills

  1. Published by Antigravity Skills · Business data analysis

    Analyzes data for business decisions: statistics, visualizations, trend detection, attribution, A/B tests, segmentation and executive reporting.

    Use it when: You have data and a business question, and need an analysis you can present.

  2. Published by Antigravity Skills · Slow SQL queries

    Debugs slow SQL with EXPLAIN analysis, designs indexes, fixes N+1 queries and rewrites queries for performance.

    Use it when: A query or page is slow and you suspect the database.

  3. Published by Antigravity Skills · Spreadsheets and financial models

    Creates, edits and analyzes spreadsheets (.xlsx, .xlsm, .csv, .tsv), builds formula-based models, preserves formatting and recalculates with error detection.

    Use it when: Your data or model lives in Excel and you want the agent to work with it without breaking formatting.

  4. Published by Antigravity Skills · Valid A/B experiments

    A structured guide for valid A/B tests: blocks invalid hypotheses, enforces statistical power and prevents peeking.

    Use it when: You are about to run an experiment and want it set up so the result means something.

  5. Published by Antigravity Skills · Graph and network analysis

    Analyzes graphs and networks in Python with NetworkX, including algorithms such as PageRank.

    Use it when: Your data is a network of relationships and you need graph algorithms.

  6. Published by Antigravity Skills · Exact symbolic math

    Symbolic math with SymPy: solve equations, do calculus and handle 2D and 3D geometry with exact results.

    Use it when: You need exact answers, not floating-point approximations.

  7. Published by Antigravity Skills · Managing Hub models and datasets

    Uses the Hugging Face Hub CLI (hf) to download, upload and manage models, datasets and Spaces.

    Use it when: You move models or datasets to and from the Hugging Face Hub.

  8. Published by Antigravity Skills · Exploring Hub datasets

    Queries, filters and paginates Hugging Face datasets through the Dataset Viewer API and returns parquet links.

    Use it when: You want to look inside a dataset on the Hub without downloading all of it.

  9. Published by Antigravity Skills · Fine-tuning language models

    Fine-tunes models with TRL methods such as SFT, DPO and GRPO on cloud GPUs, with GGUF export for local use.

    Use it when: You want to fine-tune an open model without managing GPU machines yourself.

  10. Published by Antigravity Skills · Cloud compute without setup

    Runs workloads such as data processing, batch inference and training on managed CPUs, GPUs or TPUs, and saves results to the Hub.

    Use it when: A job is too heavy for your laptop and you do not want to set up cloud machines.

  11. Published by Antigravity Skills · Experiment tracking

    Tracks ML experiments with Trackio: Python logging, alerts and metric retrieval from the CLI.

    Use it when: You are training models and need to compare runs and get alerted when something goes wrong.

  12. Published by Antigravity Skills · ML inference in JavaScript

    Runs Hugging Face models in JavaScript or TypeScript with Transformers.js, in Node.js or the browser.

    Use it when: You want model inference in a web or Node app without a Python backend.

How to install these skills

Put each skill folder (the one that contains SKILL.md) in your agent’s skills folder: .agents/skills/ in your project for Antigravity and Codex, or .claude/skills/ for Claude Code. The install guide covers global folders, CLI commands and a full worked example.

Read the install guide

Frequently asked questions

Which data skill should I start with?
Data Analyst for business questions, SQL Optimization when the database is slow, and Excel Master when the data lives in spreadsheets.
Do the Hugging Face skills need a Hugging Face account?
The skills that upload to the Hub, run Jobs or train on cloud GPUs work against your own Hugging Face account. Check each skill page for what it needs.
Do these skills work in Antigravity and Codex too?
Yes. They use the SKILL.md format, which Claude Code, Antigravity and Codex all read. Only the install folder differs.
How do I install these skills?
Download the skill and put its folder (the one that contains SKILL.md) in your agent’s skills directory: .claude/skills/ for Claude Code, or .agents/skills/ in your project for Antigravity and Codex. The install guide at /docs/install-skills lists global folders and CLI options.
What does a download cost?
Each packaged skill download on Antigravity Skills uses 1 credit. The skill pages and the install guide are free to read.
Data Science and ML Skills for Claude Code and Antigravity