Research Skills

Research Skills is a cross-agent plugin marketplace for academic research and development workflows. It ships skills, agent-callable capabilities that auto-trigger from natural-language intent, covering the parts of a research group's work that AI coding agents are good at automating: literature search, grant writing, manuscript preparation, publication figures, presentations, project lifecycle management, neuroinformatics data standards, and cloud GPU provisioning for model training.

It targets three agent runtimes from one shared set of skill definitions: Claude Code, OpenAI Codex, and GitHub Copilot CLI.

Learn it hands-on

This marketplace is taught week-by-week in the free Agentic Research Course from the Open Science Collective.

Plugins

Plugin	What it does
Project	Epics/worktrees, PR review, CI/CD, debugging, agent fan-out
Grant	NIH/NSF proposal writing, review, figure QA
Manuscript	Literature review, peer review, writing, journal formatting
Opencite	Literature search, citation management, PDF retrieval
Figures	Publication-quality figure composition and QA
Presentation	Interactive Reveal.js slide decks
Neuroinformatics	BIDS conversion, HED annotation, experiment design
ML Training	Cloud GPU provisioning, prebaked images, cost control

Start with Getting Started to install a plugin, or Architecture to see how skills, commands, agents, and MCP servers fit together.