com um clique
FortranSkills
FortranSkills contém 10 skills coletadas de aledinola, com cobertura ocupacional por repositório e páginas de detalhe dentro do site.
Skills neste repositório
Use when working on Alessandro's Windows machine through Codex, especially for PowerShell vs cmd.exe behavior, PATH refreshes, quoted Program Files paths, sandbox escalation, local command-line tools, Git/GitHub CLI, or Windows-specific setup quirks.
Proofread, analyze, and safely modify Microsoft Word .docx documents, especially academic reflective writing, application forms, assignments, and templates with section word or character limits. Use when Codex needs to extract Word document text, read surrounding instruction files for context, count words by section, check author-date references against in-text citations, verify that cited references exist, produce a proofread copy, or apply paragraph-level edits while preserving the original document.
Draft referee reports for economics and closely related academic papers from PDF manuscripts. Use when Codex needs to read a paper PDF, extract text with pdfplumber, optionally normalize extracted notes with pandoc, and produce a referee report with three sections: (1) overview and overall assessment, (2) major comments, and (3) minor comments.
Create, edit, review, build, and debug modern Fortran code and projects. Use when Codex needs to work with `.f90`, `.f95`, `.f03`, `.f08`, `.F90`, Makefiles or build scripts for Fortran, especially when the preferred compiler is Intel `ifx` from oneAPI on Windows, and when code should follow modern Fortran best practices from Fortran-lang.
Create, edit, and compile LaTeX documents. Use when working with .tex files, fixing LaTeX compilation errors, or preparing TeX sources for reliable builds, including Scientific WorkPlace (SWP) compatibility issues.
Use only for MATLAB code involving GPU-specific constructs or GPU execution, such as gpuArray, arrayfun on GPU, GPU vectorization, GPU memory use, or value function iteration intended to run on the GPU. For ordinary CPU MATLAB .m editing, use the matlab skill instead.
Use for all MATLAB .m editing, review, debugging, refactoring, and CPU performance work. This is the default MATLAB skill for scripts, functions, numerical workflows, interpolation, search, optimization, and value function iteration unless the user specifically needs GPU constructs such as gpuArray or arrayfun on GPU; in that case also use matlab-gpu.
Read and extract information from PDF documents efficiently on a Windows machine. Use when Codex needs to inspect PDFs, extract text or tables, search papers/manuals/reports, read metadata or page ranges, render pages for visual checks, repair or split difficult PDFs, or decide whether OCR is needed. Prefer Program Files tools such as Poppler, Pandoc, qpdf, and Tesseract, and check exact paths before use.
Work with Stata projects, `.do` files, logs, help files, and bundled manuals. Use when Codex needs to create or edit Stata scripts, run Stata from Windows PowerShell, inspect `.dta`-based workflows, or locate documentation shipped with the local Stata installation.
Structure literature searches and produce paper summaries and literature syntheses for economics topics. Use when Codex needs to map a research area, find recent and high-quality economics papers, summarize individual papers, compare findings across studies, or identify gaps in the economics literature using reputable sources such as JSTOR, EconLit, Google Scholar, RePEc/IDEAS, and NBER.