What problem does it solve? Research projects in hydrology, energy systems, and extremes analysis often mix R, Python, Fortran, Quarto, and LaTeX without consistent style rules or reproducibility safeguards. This Skill loads the right conventions for each file type and runs deterministic pass/fail checks only on content that actually changed, keeping research code portable and reproducible without redundant validation runs. ## Core Features & Use Cases - File-type conventions: Loads matching rule files for .R, .py, .f90, .qmd, .tex, .stan, and native build files before editing, deferring to existing local style on conflict. - Selective constraint checks: Runs check-all.py against changed paths to enforce no absolute paths, deterministic random seeds, pixi R-task dollar-sign safety, and Plotly layout text separation. - Repository hygiene: Prevents committing OS debris, editor state, and secret-bearing files like .env or .Renviron, and keeps .gitignore coverage correct. - Use Case: While editing a Fortran simulation and a Quarto paper in the same project, the Skill applies implicit none and explicit-kind conventions to the Fortran edit, then runs the seed and absolute-path checks only on the files you changed. ## Quick Start Ask the assistant to review or edit your R, Python, Fortran, Quarto, or LaTeX research files using the project conventions and run the applicable reproducibility checks on the changed files.
