Borrowing it
Nothing to install: this file belongs to saski/arnesto. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/saski/arnesto/main/.agents/skills/ml-best-practices/SKILL.mdgit clone --depth 1 https://github.com/saski/arnestoWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/saski/arnesto/ml-best-practices)<a href="https://agentmods.dev/skills/saski/arnesto/ml-best-practices"><img src="https://agentmods.dev/badge/skills/saski/arnesto/ml-best-practices.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00129 | $0.02023 |
| Opus 5 | $0.00064 | $0.01012 |
| Sonnet 5 | $0.00026 | $0.00405 |
| Haiku 4.5 | $0.00013 | $0.00202 |
Grade A, and why
ml-best-practices scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
This is a copy
100% identical to ml-best-practices — 0 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 213 lines — stays where its author put it; the contents beside it link to each section on GitHub.
ML Best Practices
I want to read a story about the data, not just run code. Ensure every code cell is followed by a markdown cell analyzing the results. End the notebook with a summary comprehensively answering the prompt.
If there is a good match between the user's request and a corresponding example plan, then adapt the example plan to fully answer the user's request:
Clustering:
Identify distinct groups based on their features.
- Understand the schema and field descriptions.
- Visualize features referenced in the prompt (e.g., with histograms, scatterplots).
- Transform dates into timestamps.
- Before applying encoders, check if the dataset already contains pre-encoded features and prefer existing numerical representations.
- Prefer to keep data instead of dropping it when possible.
- Transform ordinal data with an ordinal encoder.
- Transform nominal data with a one hot encoder.
- Standardize numerical features.
- Perform clustering with a range of values, and collect the silhouette score.
- Choose the optimal number of clusters based on the silhouette score.
- Use dimensionality reduction (e.g., PCA) to project the data into two dimensions.
- Scatterplot the samples in two dimensions with cluster labels as the hue.
- Scatterplot the samples in two dimensions with a discrete feature as the hue.
- Describe the clusters in text by feature distributions or typical feature values.
- Conclusion: comprehensively answer the prompt in a final markdown cell.
Time Series Forecasting:
Develop a predictive model to estimate future values based on historical trends. How might different modeling approaches impact the prediction accuracy?
- Understand the schema and field descriptions.
- Visualize the target feature over time at a reasonable granularity.
- Always perform a chronological split on the data to create training, validation, and test sets.
- Are there seasonal trends?
- Test for stationarity.
- Discuss possible modeling approaches. How might different modeling approaches impact the prediction accuracy?
- Train two time series forecasting models to predict the target feature. Use previous seasonality and stationarity information as model hyperparameters.
- Predict the target feature for the training and validation sets.
- Optionally, hypertune models with the validation set.
- Visualize the actual and predicted target feature vs time for each model on the training and validation sets.
- Evaluate the validation performance with error metrics.
- Select a model.
- Retrain the selected model on the test and validation sets.
- Predict the test values with the selected model.
- Visualize the average target feature and the predicted test values.
- Conclusion: comprehensively answer the prompt in a final markdown cell.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 3d ago First seen · 213 lines · 129 tokens per session scan A ae010fd52e6b
ml-best-practices is a skill published in the GitHub repository saski/arnesto (5 stars, last pushed today), licensed Unlicense. It adds 129 tokens to every session and 2,023 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it A with 0 findings. It is 100% identical to ml-best-practices, differing in 0 lines, and is treated as a copy.
Other skills, from other repositories
digital-health-clinical-asr-finetune
Stage 4 of the Clinical ASR Flywheel. Use when priority KER is above 0.3 to run stock NeMo SFT on Parakeet TDT v2 and offline cycle N+1 re-eval. NOT for generic word boosting (use /finetune-asr).
digital-health-clinical-asr-eval
Stage 3 of Clinical ASR Flywheel. Score a NeMo manifest, produce the five-section KER leaderboard (by-ipasource diagnostic). Not for ASR auth (/riva-asr).
i4h-workflow-dataset-annotate
Grade or filter workflow HDF5 episodes with an OpenAI-compatible vision model. Use for visual success labels; do not use for replay, policy evaluation, or recordings without frames.
i4h-workflow-dataset-convert
Convert workflow HDF5 recordings to LeRobot datasets for training or browser inspection. Use for conversion; do not use for replay, augmentation, or raw-data repair.
i4h-lerobot-viz
Serve and visually inspect a converted LeRobot dataset in the browser. Use for videos and state/action timelines; do not use for raw workflow HDF5 or incomplete conversion output.
i4h-workflow-dataset-mimic
Expand workflow HDF5 demonstrations with action jitter, optionally scoped to node segments. Use for synthetic variants; do not use to collect data, alter state directly, or generate new images.