/README

Multi-Agent OS Commands

Overview

Slash commands for the multi-agent-os plugin. Commands are auto-discovered by Claude Code from this directory.

Available Commands

Command	File	Description
/sync	sync.md	Sync content from framework to consumer
/audit	audit.md	On-demand session/agent/task auditing
/agentic-status	agentic-status.md	Display human-readable status maps
/worktree	worktree.md	Manage git worktrees
/delegate	delegate.md	Delegate tasks to sub-agents
/founder-playbook	founder-playbook.md	Diagnose AI-native startup stage + route to stage skills

Command Categories

Synchronization

• /sync — Keep consumer projects updated with framework

Observability

• /audit — Deep analysis via Sentinel Protocol
• /agentic-status — Quick visual status check

Coordination

• /worktree — Git worktree management
• /delegate — Sub-agent task delegation

Startup Coaching

• /founder-playbook — Diagnose AI-native startup stage; route to founder-stage-* skills

Command Structure

Each command file uses frontmatter:

---
name: command-name
description: Brief description
---

# /command-name Command

[Documentation...]

Usage

Commands are invoked in Claude Code conversations:

/sync all
/audit session
/agentic-status
/worktree create c614-feature
/delegate analyst "Review the requirements"

Adding New Commands

1. Create {name}.md in this directory
2. Add frontmatter with name and description
3. Document usage, options, and examples
4. Commands are auto-discovered on plugin load

---

Part of multi-agent-os plugin v1.0.0