Skip to main content
affpapa.org
RU EN
skill

aeon — Claude Code skill: time series machine learning

Gives Claude a working map of the aeon Python library for classification, forecasting, clustering and anomaly detection on time series.

Install into Claude Code · claude-code-templates All methods and platforms ↓ 
npx claude-code-templates@latest --skill scientific/aeon
Open on GitHub

What aeon does

This skill teaches the agent how to run time series machine learning with aeon. It covers classification, regression, clustering, forecasting, anomaly detection, segmentation and similarity search, each with code samples and notes on picking an algorithm. Want speed? MiniRocketClassifier or Arsenal. Want accuracy? HIVECOTEV2 and InceptionTimeClassifier. Need to explain a decision? ShapeletTransformClassifier and Catch22Classifier.

There are also sections on feature extraction with RocketTransformer and Catch22, on temporal distance measures such as DTW, and on deep architectures including FCNClassifier and ResNetClassifier. Standard benchmark datasets load straight from the package, and get_estimator_results pulls published scores for comparison. The API follows scikit-learn, so extracted features drop into any sklearn estimator.

Install the library first with uv pip install aeon — the snippets will not run without it. Some algorithms get slow on large collections, and the deep learning networks ask for more compute than the classical ones.

What it helps with

  • label time series into classes and score several classifiers
  • forecast the next values of a series with ARIMA
  • group unlabeled series by shape
  • flag anomalies and change points in sensor data
  • turn a series into features for a sklearn model

Example prompts for the agent

  • Pick a time series classifier for the GunPoint dataset and report test accuracy
  • Find recurring motifs in sensor.csv with StompMotif
  • Compare DTW against Euclidean distance on my series

How to install aeon

1claude-code-templates

npx claude-code-templates@latest --skill scientific/aeon

2Skills CLI (Vercel)

npx skills add davila7/claude-code-templates --skill aeon -a claude-code

3Manually: put the files in place

~/.claude/skills/aeon/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/aeon/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about aeon

How do I install aeon in Claude Code?

Run npx claude-code-templates@latest --skill scientific/aeon. Or use npx skills add davila7/claude-code-templates --skill aeon -a claude-code.

What does the aeon skill do?

It hands the agent ready recipes for time series ML: classification, regression, clustering, forecasting, anomaly detection, segmentation and similarity search.

What do I need before using it?

The aeon library itself, installed with uv pip install aeon. The examples rely on its scikit-learn compatible API.

Similar skills

All skills for AI and ML →
Time Series Analysis aj-geddes A skill that walks your agent through time series work: decomposition, stationarity checks, ARIMA and forecasts in Python. Useful for sales or traffic data. CX Codex +1 605installs alphaear-predictor RKiding This skill produces a Kronos-based forecast for a financial time series, then adjusts the numbers against fresh news. Useful when you need a short-horizon price view. CX Codex +1 584installs scikit-learn K-Dense-AI A skill for Python machine learning tasks: classification, regression, clustering, preprocessing, and building ML pipelines with scikit-learn 1.9.1. CX Codex +1 2,200installs scikit-survival K-Dense-AI A skill that walks an agent through scikit-survival 0.28.0: building, tuning and auditing survival models on censored data, with metrics that account for censoring. CX Codex +1 2,000installs Anomaly Detection aj-geddes Detect outliers, fraud, and system faults using statistical tests, isolation forests, and deep learning. Generates working Python analysis pipelines. CX Codex +1 523installs data-scientist VoltAgent Community Subagent for exploratory data analysis, hypothesis testing, and building machine learning models directly in Claude Code. 3,342installs

From the same set495

The whole set →
interview-prep-generator Turns your resume into STAR stories, likely questions and talking points for a specific role. Handy before a job interview. 61installs gws-modelarmor-create-template A slash command that has Claude create a Google Model Armor template in GCP, either from the jailbreak preset or from your own JSON filter config. 61installs emerging-techniques-speculative-decoding A skill that walks your agent through speculative decoding, Medusa and lookahead decoding to speed up token generation by 1.5–3.6×. 59installs claude-opus-4-5-migration A skill that scans your codebase for model strings and API calls and switches them to Opus 4.5. Useful when a project needs the newer model. 55installs upstash-qstash A skill about Upstash QStash: delayed HTTP calls, cron jobs and webhook delivery. Useful when you want an HTTP request to fire later and actually arrive. 51installs sveltekit A Claude skill that guides work on SvelteKit apps: routes from files, server rendering, endpoints and form handling. 50installs regression-triage A slash command that walks a regression backwards: reruns the failing test, reads history around the broken path, and names the commit that did it. 25installs doordash-group-checkout-gate A hook that re-checks allergens at DoorDash group checkout through dd-cli and blocks payment when a dangerous match turns up. 24installs

Project sponsors

Companies that keep this analytics open
Gold sponsor
Nutra.Club

Impressions, clicks and code copies — open statistics. We keep the numbers public for the same reason NeBlask itself is public. Become a sponsor