Omt

OMT - One Man Team Plugin

A streamlined Agent-First development workflow for Claude Code. Your personal development squad that makes one person feel like a team.

Overview

OMT (One Man Team) transforms your development workflow:

• One command — /omt "goal" launches the full lifecycle
• One decision point — approve the consensus, then sit back
• Agents execute — autonomous development until completion or escalation
• Contracts connect — clear input/output definitions between agents

Hive Mode: Autonomous Lifecycle

The /omt command launches @hive, the lifecycle coordinator that handles everything:

/omt "Build user authentication with JWT"
  │
  ▼
@hive initializes workspace
  │
  ▼
@hive ──dispatch──▶ @pm (autonomous requirements)
                    └─▶ .agents/outputs/pm.md
  │
  ▼
@hive ──dispatch──▶ @arch (autonomous architecture)
                    └─▶ .agents/outputs/arch.md
  │
  ▼
@hive presents consensus summary
  + all decision points collected upfront
  │
  ▼
Human approves (single interaction)
  │
  ▼
For each task:
  @hive ──dispatch──▶ @dev (TDD implementation)
  @hive ──dispatch──▶ @reviewer (review + commit)
  (retry ≤3, then escalate)
  │
  ▼
Completion report → .agents/outputs/hive.md

Key Design: All decision points are front-loaded into the consensus gate. After approval, execution proceeds without interruption.

Features

5 Core Agents

Agent	Phase	Model	Purpose
@hive	Coordination	claude-sonnet-4-5	Full lifecycle coordinator — dispatches all other agents
@pm	Planning	claude-haiku-4-5	Requirements management and clarification
@arch	Planning	claude-sonnet-4-5	API-First architecture design
@dev	Execution	claude-sonnet-4-5	Development implementation (TDD + debugging)
@reviewer	Review	claude-sonnet-4-5	Code review + git commit authority

Commands

Command	Purpose
/omt <goal>	Launch autonomous lifecycle — the primary entry point
/init-agents	Initialize agent workspace (once per project)
/help	Help and command reference
/git-commit	Emergency manual commit

Contract-First Design

Planning Phase Outputs (Autonomous via @hive):

Agent	Output	Contract
Human	goal.md	-
@pm	outputs/pm.md	pm.json
@arch	outputs/arch.md	arch.json

Execution Phase Contracts:

Contract	Connection	Definition
dev.json	@dev → @reviewer	Implementation results and test coverage
hive.json	@hive lifecycle	Full lifecycle coordination

Skills:

• contract-validation - Validate agent contracts

Git Workflow

Commit Authority:

Has commit authority:

• @reviewer (automatic after review)
• /git-commit (manual, emergency only)

No commit authority:

• All other agents (including @hive)

Commit Format:

<type>[optional scope]: <description>

Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

Types: feat, fix, docs, refactor, test, chore

Failure Protection

• Auto-escalation after 3 consecutive failures
• Clear error reporting with actionable suggestions
• All decision points front-loaded to prevent mid-execution blocks

Installation

1. Add Marketplace

/plugin marketplace add musingfox/cc-plugins

2. Install Plugin

/plugin install omt

3. Initialize Workspace

In your project root:

/init-agents

This will:

• Create .agents/ directory structure with .state/ infrastructure
• Configure task management system
• Update .gitignore rules

Usage

Quick Start

# 1. Initialize workspace (once per project)
/init-agents

# 2. Launch autonomous workflow
/omt "Implement a REST API for user management with CRUD operations"

# 3. Review the consensus summary when @hive presents it
# 4. Approve → autonomous execution begins
# 5. Done!

Example: New Feature

# Launch with a detailed goal
/omt "Build user authentication with JWT, including login, signup, password reset, and role-based access control"

# @hive will:
# 1. Dispatch @pm to define requirements autonomously
# 2. Dispatch @arch to design architecture autonomously
# 3. Present a consensus summary with all decision points
# 4. After your approval, execute @dev → @reviewer for each task
# 5. Generate a completion report

Agent Workspace

Structure

.agents/
├── .gitignore           # Ignores .state/
├── goal.md              # Human's goal
├── outputs/
│   ├── pm.md            # @pm requirements
│   ├── arch.md          # @arch architecture
│   ├── dev/             # Per-stage @dev reports
│   ├── reviews/        # Per-stage @reviewer reports
│   └── hive.md          # @hive completion report
└── .state/              # Infrastructure (gitignored)
    ├── config.json      # Workspace configuration
    ├── state.json       # Task state
    ├── hive-state.json  # @hive lifecycle state
    └── tasks/           # Task tracking data

Library (lib/)

• contract-validator.ts - Contract validation
• state-manager.ts - State management

Documentation

• Quick Start: See commands/help.md
• Command Reference: See commands/ directory
• Agent Specifications: See agents/ directory
• Contracts: See contracts/ directory

License

MIT License

Support

For issues and feedback:

• GitHub Issues: https://github.com/musingfox/cc-plugins/issues
• Documentation: Plugin installation includes complete docs

---

Last Updated: 2026-03-01 Status: Production Ready