pegasus-ai

SciTech Claude Code Plugin Marketplace

A curated collection of Claude Code plugins for scientific computing — built for researchers and developers working with Pegasus WMS and SciTech infrastructure.

What Is This?

Claude Code plugins extend the Claude Code CLI with domain-specific skills, MCP servers, and AI-assisted workflows. This marketplace provides plugins tailored for scientific computing — including Pegasus WMS workflow authoring and SciTech project development.

Prerequisites

• Claude Code CLI installed

Quick Start

Add the marketplace and install a plugin in two commands:

/plugin marketplace add pegasus-isi/claude-plugin-marketplace
/plugin install <plugin-name>@scitech

Installation

Option 1: Claude Code CLI (recommended)

Run the following commands from within Claude Code:

/plugin marketplace add pegasus-isi/claude-plugin-marketplace
/plugin install <plugin-name>@scitech

Option 2: Manual configuration

Add the following to your .claude/settings.json file:

{
  "extraKnownMarketplaces": {
    "scitech": {
      "source": {
        "source": "github",
        "repo": "pegasus-isi/claude-plugin-marketplace"
      }
    }
  },
  "enabledPlugins": {
    "<plugin-name>@scitech": true
  }
}

Available Plugins

Plugin	Description
pegasus-dev	Skills and tools for developing software on SciTech projects — git workflows, code review, commit conventions, and project-specific best practices.
pegasus-ai	Workflow authoring for Pegasus WMS — generate workflow.yml files, experiment configs, and scaffold scientific pipelines with Claude.
impeccable	Design vocabulary and skills for frontend development. Includes 20 commands (/polish, /distill, /audit, /typeset, /overdrive, etc.) and an enhanced frontend-design skill with curated anti-patterns.
nano-banana	Nano Banana image generation supporting Gemini Flash, Pro, and Nano Banana 2 models.

Contributing

See CONTRIBUTING.md for instructions on how to add or update plugins.

Funding

Funded by National Science Foundation (NSF) under award 2513101.

License

Apache 2.0 © Pegasus ISI