oh-my-claudecode
github.com/yeachan-heo/oh-my-claudecode- Category
- AI Agents
- Rank
- No. 101Tools index
- Pricing
- Open Source
- Type
- TOOL
- Builder
- yeachan-heo
- GitHub
- 39.1k stars
- Latest release
- v5.3.0
- Date
About
Multi-agent orchestration framework for Claude Code that enables teams of specialized AI agents to work together on complex coding tasks. Provides zero-configuration setup with natural language commands and automatic task delegation across 32+ specialized agents.
What it does
This is a Claude Code add-on with guided build modes, persistent task state, verification loops, background work, and a live status display. Users can work through interactive session commands, launch terminal workers in tmux, or assemble local Node.js programs with its TypeScript helpers.
Why it's ranked here
The breadth is real, not just prompt packaging. The source exposes task decomposition, model routing, recovery, context injection, verification, worker health, permissions, worktrees, and merge coordination. That depth makes it compelling for complex repository work, although its split runtimes and platform constraints increase operational complexity.
What's good
It supports several levels of control without collapsing them into one interface. Interactive workflows handle open-ended coding, terminal commands suit repeatable automation, and the library surface supports custom Agent SDK programs. Team operations include heartbeats, task routing, recovery checkpoints, permission rules, audit logs, usage reports, and merge coordination.
Tradeoffs
The feature set has a large conceptual surface, with distinct plugin, terminal, library, and team runtimes. Interactive workflows are unsuitable for headless CI. Terminal teams depend on tmux, native Windows receives a warning, and named workflow profiles currently require Linux plus flock. The npm installation may also emit an upstream native dependency deprecation warning.
How to use it well
Use it when Claude Code is already central to your workflow and larger tasks benefit from planning, parallel execution, recovery, and explicit verification. Prefer terminal commands for deterministic CI tasks and interactive modes for supervised development. It does not provide a VS Code extension, and its library API does not replace the Claude Code plugin interface.
Technical notes+
package.json defines an ESM npm package requiring Node.js 20 or newer, with omc, oh-my-claudecode, and omc-cli binaries, a TypeScript build, Vitest scripts, and dependencies on the Claude Agent SDK, MCP SDK, SQLite, Zod, and Commander. src/index.ts builds Agent SDK query options, loads project context, configures agents and MCP servers, selects allowed tools, processes trigger keywords, and manages background tasks. src/cli/index.ts implements the terminal surface and warns on native Windows. src/team/index.ts exposes tmux sessions, task files, messaging, heartbeats, recovery, worktrees, routing, permissions, scaling, and merge coordination. src/hud/index.ts reads Claude Code status input and renders session state. src/mcp/index.ts and src/tools/index.ts expose MCP configuration plus LSP, AST, and Python tools.
Observed
- License
- MIT
- Primary language
- TypeScript
- Runtime
- Node.js 20 or newer, using ECMAScript modules
- Install surface
- Claude Code marketplace plugin or global npm package named oh-my-claude-sisyphus
- Interfaces
- Claude Code in-session skills, terminal CLI, TypeScript library, and MCP tooling
- Platform constraints
- tmux is required for CLI teams; native Windows is warned; named workflow profiles require Linux and flock
- Testing
- Vitest test, run, UI, and coverage scripts are declared in package.json
Read from README.md, package.json, src/index.ts, src/cli/index.ts, src/hud/index.ts, src/mcp/index.ts, src/team/index.ts, src/hooks/index.ts, src/tools/index.ts, src/agents/index.ts, src/config/index.ts, src/shared/index.ts, src/commands/index.ts, src/features/index.ts.
What it can do
Orchestrate multiple AI agents for complex coding tasks
Complex coding project requirements → Coordinated work output from specialized AI agents
Delegate tasks automatically across specialized agents
Natural language coding commands → Task assignments to appropriate specialized agents
Execute coding tasks with zero configuration setup
Natural language instructions → Working code solutions
Parallelize coding work across agent teams
Multi-part coding requirements → Parallel execution results from multiple agents
Route tasks to optimal AI models intelligently
Coding task specifications → Model-optimized task execution with reduced token usage
Coordinate team-based AI agent collaboration
Team coding objectives → Synchronized multi-agent coding workflow
Tags
Tech Stack
Media
Comments (0)
No comments yet
Editorially curated, with community endorsements as a secondary signal. Corrections welcome.