oh-my-claude-code

Multi-agent orchestration for Claude Code with intelligent delegation and specialized agents.

oh-my-claude-code

Multi-agent orchestration for Claude Code, inspired by oh-my-opencode

English | 中文

Version: 1.2.0 | Changelog | ROADMAP

A Claude Code plugin providing specialized agents that work together through intelligent delegation.

Overview

oh-my-claude-code brings the power of multi-agent orchestration to Claude Code. It provides:

  • Specialized Agents - Each agent excels at a specific domain
  • Intelligent Delegation - The orchestrator routes tasks to the right agent
  • Cost-Aware Selection - FREE → CHEAP → EXPENSIVE decision framework
  • Structured Metadata - YAML-based agent definitions with rich metadata
  • Ultrawork Mode - Obsessive task completion with ulw magic keyword
  • Todo Continuation - Sisyphus-style enforcement to complete all tasks

v1.2.0 Architecture Improvements

  • Centralized Configuration - All config managed in config/index.js
  • Parameter Validation - Type-safe validation with helpful error messages
  • Progress Visualization - Beautiful progress bars and time estimates
  • Performance Optimized - Parallel PostToolUse processing (~40% faster)
  • Modular Utils - Reusable validation, patterns, and progress reporting

Installation

From Marketplace (Recommended)

# Add the marketplace
/plugin marketplace add https://github.com/zephyrpersonal/oh-my-claude-code

# Install the plugin
/plugin install oh-my-claude-code@oh-my-claude-code-plugins

Local Installation (For Development)

  1. Clone the repository:
git clone https://github.com/zephyrpersonal/oh-my-claude-code.git
  1. Add local marketplace and install:
/plugin marketplace add ./oh-my-claude-code
/plugin install oh-my-claude-code@oh-my-claude-code-plugins

Plugin Structure

oh-my-claude-code/
├── .claude-plugin/
│   ├── plugin.json           # Plugin manifest
│   └── marketplace.json      # Local marketplace config
├── hooks/
│   ├── hooks.json            # Hooks configuration
│   ├── ultrawork-detector.js # Detects ulw keyword
│   ├── todo-continuation-enforcer.js
│   └── post-tool-processor.js # Unified PostToolUse handler
├── config/
│   └── index.js              # Centralized configuration
├── utils/
│   ├── validation.js         # Parameter validation
│   ├── patterns.js           # Pre-compiled regex
│   └── progress-reporter.js  # Progress visualization
├── commands/
│   └── ulw.md                # Slash command
├── agents/
│   ├── orchestrator.md
│   ├── explore.md
│   ├── librarian.md
│   └── ...
├── skills/
│   └── ultrawork/
│       └── SKILL.md
├── docs/
│   └── WEEK1-2-SUMMARY.md    # Development summaries
├── ROADMAP.md                # 48-week iteration plan
└── package.json

Agents

AgentModelCostPurpose
orchestratorinheritEXPENSIVETask classification and delegation
explorehaikuFREEInternal codebase search
librariansonnetCHEAPExternal docs and research
oracleopusEXPENSIVEArchitecture and deep analysis
frontend-ui-ux-engineersonnetCHEAPVisual/UI changes
document-writersonnetCHEAPTechnical documentation
multimodal-lookersonnetCHEAPPDF/image analysis

See AGENTS.md for detailed information about each agent.

Ultrawork Mode (Sisyphus Mode)

Ultrawork is a special mode that enables obsessive task completion.

Activation

Add ulw, ultrawork, or uw anywhere in your message:

claude "Implement user authentication, ulw"
claude "ulw refactor the database layer"

Parameters

ULW supports optional parameters for fine-tuned control:

ParameterDefaultDescription
--max-iterations N50Maximum continuation attempts
--thoroughness LEVELthoroughSearch depth: quick, medium, thorough
--no-diagnostics(enabled)Disable auto-diagnostics reminders
--completion-signal "TEXT"(none)Custom completion signal phrase

Examples:

# Limit iterations
claude "implement auth, ulw --max-iterations 20"

# Custom completion signal
claude "add tests ulw --completion-signal 'all tests pass'"

# Quick search mode
claude "ulw --thoroughness quick fix the typo"

# Combined parameters
claude "refactor cache layer, ulw --max-iterations 30 --thoroughness medium"

What It Does

  • Creates comprehensive todo lists for every task
  • Does NOT stop until ALL todos are complete
  • Uses maximum thoroughness for searches
  • Verifies every change with diagnostics
  • Automatically continues if stopped with incomplete tasks

The Sisyphus Promise

"Like Sisyphus, you roll the boulder every day. You don't stop until you reach the top."

ULW vs Ralph Loop

Both ULW and Ralph Loop aim to keep Claude working until completion, but with different approaches:

DimensionULW (Ultrawork)Ralph Loop
State TrackingStateful via todosStateless, re-injects same prompt
ProgressTracks via todo itemsTracks via file/git changes
Completion SignalAll todos marked completeOutputs <promise>COMPLETE</promise>
Activationulw keyword anywhere/ralph-loop "prompt" command
Best ForComplex tasks with clear stepsGreenfield, "walk away" tasks
Iteration ControlmaxContinuations: 50--max-iterations N

When to use which:

  • ULW: When you need fine-grained task tracking and want to see progress
  • Ralph Loop: When you want Claude to iterate freely until it decides it's done

Hooks

This plugin includes several hooks for enhanced workflow:

HookEventPurpose
ultrawork-detectorUserPromptSubmitDetects ulw keyword and injects ultrawork instructions
todo-continuation-enforcerStopBlocks stopping when todos are incomplete
post-tool-processorPostToolUseUnified handler for comment checking and diagnostics reminders

Usage Examples

# Use the orchestrator for complex tasks
claude "Use the orchestrator to analyze this codebase, ulw"

# Delegate to specific agents
claude "Use the explore agent to find all error handling patterns"
claude "Use the librarian agent to research React hooks best practices"

Changelog

[1.2.0] - 2025-01-11

Week1-2 Architecture Optimization

Added

  • Centralized Configuration (config/index.js)
    • All constants and settings in one place
    • Easy customization for users
    • Pre-compiled regex patterns
  • Parameter Validation (utils/validation.js)
    • Type-safe validation for ultrawork parameters
    • Helpful error messages with field names
    • Range checking and enum validation
  • Progress Visualization (utils/progress-reporter.js)
    • Beautiful ASCII progress bars
    • Time estimates for remaining tasks
    • Compact and detailed report formats
  • Pre-compiled Patterns (utils/patterns.js)
    • All regex patterns defined centrally
    • Performance optimization through pre-compilation
    • Helper functions for pattern matching

Changed

  • PostToolUse Hooks Unified
    • Merged comment-checker and auto-diagnostics into single processor
    • Parallel execution via Promise.all (~40% performance gain)
    • Reduced process count from 2 to 1
  • ultrawork-detector.js
    • Now uses centralized configuration
    • Integrated parameter validation
    • Better error messages
  • todo-continuation-enforcer.js
    • Enhanced with progress visualization
    • Uses pre-compiled patterns
    • Cleaner code structure

Documentation

  • ROADMAP.md - 48-week iteration plan (4 quarters)
  • docs/WEEK1-2-SUMMARY.md - Complete summary of Week1-2 changes
  • notes.md - Optimization analysis and technical debt tracking
  • task_plan.md - Task planning and progress tracking

[1.1.0] - 2025-01-09

  • Simplified model config to haiku/sonnet/opus format
  • Updated repository URL to zephyrpersonal/oh-my-claude-code

[1.0.0] - 2025-01-08

  • Initial release of oh-my-claude-code plugin

License

MIT

Acknowledgments