gjbex/scientific-training-development-skills
Design, review, and edit scientific-computing training material with reusable agent skills.
Create, edit, or restructure Jupyter notebooks, especially training and tutorial notebooks, while preserving a navigable outline and reliable section collapsing. Use for changes to instructional content, Markdown, code cells, exercises, or section order. This skill does not cover broader output policy, reproducible execution, CI, dependencies, data provenance, or reusable-module extraction.
Design, review, restructure, or reality-check concept-led scientific-computing training for PhD students and researchers in academia, industry, or SMEs across diverse scientific domains. Use for instructor-led, self-paced, or hybrid courses when treating presentation slides as guides for live teaching rather than standalone material, prioritizing durable principles over volatile library APIs, rejecting imitation or transcription exercises, defining measurable learning objectives, aligning content and exercises, estimating realistic scope and timing, managing cognitive load and prerequisites, balancing theory with practice, grouping optional hands-on sessions, requiring complete exercise solutions, checking opening logistics, or producing a structured evidence-based course review.