Best skills for Data scientist
For data scientists, data quality and computational scalability are non-negotiable. The AI coding skills in this directory zero in on two critical areas: historical data validation and HPC cluster management. Historical data validation tools automate the detection of inconsistencies, outliers, and sampling biases across large datasets—a must for building trustworthy models. Watch out for brittle validation pipelines: opt for smart approaches that flag issues without blindly discarding records. On the compute side, managing HPC clusters (like TACC Vista) lets you scale beyond a single workstation, accelerating not just training but also data preprocessing and hyperparameter tuning. The trick is to profile memory usage to avoid over-provisioning, and to orchestrate parallel jobs efficiently. These skills bridge the gap between raw data and production-ready insights, ensuring your models are both accurate and performant.
4 skills selected
Pointline Data Lake Research API
Data & AI
Reference for the Pointline data lake research API used by quant researchers. Covers querying event data (trades, quotes, orderbook, CN L2/L3 tables), symbol discovery/resolution, building time-series spines (clock, trades, volume, dollar bars), decoding fixed-point scaled integers, attaching PIT-correct symbol metadata, filtering by trading phase, and ingesting Bronze files into Silver. Useful when working with a PIT-accurate offline data lake built on Polars and Delta Lake.
PDF File Processing
Data & AI
Extracts text from PDFs, fills forms, and merges documents. Includes scripts for analyzing structure, extracting content, filling form fields, and validating outputs. Useful for automated PDF document processing.
Transfer Learning Model Adapter
Data & AI
Automates the adaptation of pre-trained models using transfer learning, including fine-tuning, data validation, and performance optimization. Helps when fine-tuning a model on a new dataset or adapting it for a specific task.
Model Explainability Tool
Data & AI
Automatically assists with model explainability tool operations in ML training. Provides step-by-step guidance, generates production-ready code and configurations, and validates outputs against common standards.
How to choose
- How are these skills selected?
- Each skill is curated and verified by the Skills Guides editorial team. We run a security and quality review on every entry, so only verified skills appear in this selection.
- What do the security ratings mean?
- We label skills Safe, Caution or Risky based on our security analysis — checking for prompt-injection risks, requested permissions and other red flags. The rating gives you an at-a-glance sense of how much trust a skill warrants.
- How do I install a skill?
- Open any skill page and follow its install instructions for your tool — Claude Code, Cursor or Copilot. Each skill lists the exact steps so you can get it running in a couple of minutes.