Concepts

Component Types

orbitant-os organizes capabilities into three component types:

Skills

Instructions Claude loads automatically when relevant. Claude reads the description field and decides when to activate them.

Agents

Specialized sub-agents with restricted tool access. Invoked via slash command or by another agent.

Commands

Slash commands invoked manually by the user via /plugin-name:command-name.

Comparison

Component	What it is	How it triggers
Skills	Instructions Claude loads automatically when relevant	Claude reads the `description` and decides
Agents	Specialized sub-agents with restricted tools	Invoked via slash command or by another agent
Commands	Slash commands invoked manually	User types `/orbitant-{vertical}:command-name`

Why Verticals?

Each vertical is an independent plugin with its own namespace. This means:

No collisions: Skills use prefixed names (e.g., orbitant-blog-post-review) so Claude can distinguish them from community skills.
Install only what you need: A marketer doesn’t need infra skills, and vice versa.

File Structure

Each plugin follows this structure:

plugins/orbitant-{vertical}/
├── .claude-plugin/
│   └── plugin.json          # Plugin metadata + version
├── skills/
│   └── {skill-name}/
│       ├── SKILL.md         # Required (frontmatter + instructions)
│       ├── README.md        # Optional (human docs)
│       ├── scripts/         # Optional (executable code)
│       ├── references/      # Optional (docs loaded on demand)
│       └── assets/          # Optional (templates, fonts, images)
├── agents/
│   └── {agent-name}.md      # Specialized sub-agents
└── commands/
    └── {command-name}.md    # Slash commands

Naming Conventions

Skill names: orbitant-{short-description} (e.g., orbitant-blog-post-review)
Skill folders: kebab-case (e.g., blog-post-review/)
Agent files: kebab-case .md (e.g., content-strategist.md)
Command files: kebab-case .md (e.g., generate-copy.md)
Plugin folders: orbitant-{vertical} (e.g., orbitant-marketing)