hughyau/academicforge

Academic Forge is a curated library of AI skill add-ons for researchers who use AI coding assistants in their scientific work. If you use tools like Claude Code, OpenCode, or Codex to help write code or run analyses, Academic Forge lets you install pre-built skill packs that teach those assistants how to handle specific research tasks: predicting protein structures, reviewing literature, running genomics pipelines, making publication-quality figures, and more. Think of it like a plug-in manager for your AI assistant. Instead of explaining from scratch what AlphaFold2 or scGPT does every time you start a session, you install the relevant skill and the assistant already knows the workflow, the tools involved, and the expected outputs. You browse a website, check the boxes for the skills you want, and it generates a one-line install command. You run that command in your project folder and the skills are ready to use. The library currently includes 32 skills organized under six research themes. The structure prediction group covers tools like AlphaFold2, Boltz, and DiffDock for predicting how proteins fold and how small molecules bind to them. The genomics group includes tools for single-cell RNA sequencing analysis and genome-wide function prediction from DNA sequences. A separate group handles protein design, helping you generate amino acid sequences for a given protein backbone. Figure and visualization skills help produce charts at the standard expected for journal submissions. Literature and writing skills handle tasks like synthesizing papers, exploring PDFs, and structuring the narrative of a research manuscript. A final group covers compute workflows: setting up remote GPU environments, submitting SLURM jobs, and managing model endpoints. Installation works on macOS, Linux, and Windows via either a bash or PowerShell script. The repository and the website tooling are MIT licensed. Third-party skills bundled from other projects keep their original licenses.