</>MCP Agents Market
Plugin

oh-my-openagent

by code-yeongyu69.4kTypeScriptUpdated 2026-09-24

omo/lazycodex: The coding agent for tokenmaxxers;the one and only agent harness for complex codebases. For your Codex, for your OpenCode

Claude CodeOpenCodeCodex

oh-my-openagent is a comprehensive Claude Code plugin that transforms OpenCode into a multi-agent orchestration platform with advanced coding capabilities. It provides 11 specialized AI agents (orchestrator, planner, architect, librarian, etc.) working in parallel, hash-anchored edits for surgical code changes, LSP integration, and built-in MCP servers for web search and documentation. The plugin supports three editions: Ultimate (full-featured OpenCode plugin), Light (Codex CLI components), and Senpi (standalone beta). It eliminates manual model juggling by automatically routing tasks to appropriate models based on work category, with features like Team Mode for coordinated multi-agent collaboration and ultrawork mode for end-to-end autonomous task execution.

Key Features

11 discipline-specialized agents with automatic category-to-model routing (visual-engineering, ultrabrain, deep, quick)
Hash-anchored edit tool with LINE#ID references to prevent stale-line errors (opt-in hashline_edit mode)
Team Mode with up to 8 parallel agents, real-time tmux visualization, and dedicated collaboration tools
Ultrawork mode (ulw) for autonomous end-to-end task execution with interview-based strategic planning
Built-in MCP servers for websearch (Exa), context7 (docs), grep_app (GitHub search), and LSP
Full LSP integration with diagnostics, navigation, symbols, and workspace-wide renaming
AST-Grep pattern-aware code search and rewriting across 25 languages
Automatic AGENTS.md hierarchical context injection and deep initialization (/init-deep)

Use Cases

  • 01Orchestrating multiple AI models (Claude, GPT, Gemini, Kimi) for complex multi-file refactoring tasks
  • 02Running parallel security audits with specialized vulnerability hunters and PoC engineers (security-research skill)
  • 03Converting large codebases (45k+ lines) between frameworks with autonomous overnight execution
  • 04Planning and executing complex features through interview-based strategy sessions before coding
  • 05Maintaining zero-slop code quality with comment checking and architectural consistency enforcement
  • 06Debugging production issues with architect consult agent and interactive tmux REPL sessions

Related Plugins

View more

oh-my-openagent — FAQ

What is oh-my-openagent and how does it differ from vanilla Claude Code?+

oh-my-openagent is a plugin that extends OpenCode with multi-agent orchestration, allowing 11 specialized agents to work in parallel with automatic model routing. Unlike vanilla Claude Code, it provides hash-anchored edits to prevent stale-line errors, built-in MCP servers, LSP integration, and autonomous ultrawork mode for end-to-end task execution without manual intervention.

How do I install oh-my-openagent?+

For the Ultimate edition (OpenCode), run 'bunx oh-my-openagent install' and let an LLM agent walk through the TUI setup. For the Light edition (Codex CLI), use 'npx lazycodex-ai install'. For the standalone Senpi beta, install with 'npm i -g omo-ai@beta'. The installer handles subscription detection, model configuration, and provider authentication automatically.

Which AI clients and models does oh-my-openagent support?+

oh-my-openagent works with OpenCode (Ultimate edition), Codex CLI (Light edition), and as a standalone tool (Senpi beta). It supports Claude (Anthropic), GPT models (OpenAI/Copilot), Gemini (Google), Kimi K3, GLM/Z.ai, and others through automatic provider routing. The plugin is fully compatible with Claude Code hooks, commands, skills, MCPs, and plugins.

Do I need API keys or subscriptions to use oh-my-openagent?+

Yes, you need active subscriptions or API access to at least one supported provider (Anthropic, OpenAI, Google, etc.). The installer includes a subscription interview to detect what you have. Recommended entry-level setup: ChatGPT subscription ($20), Kimi Code ($19), and GLM Coding Plan ($10) for full ultrawork functionality.

Is oh-my-openagent free and open source?+

The plugin itself is open source under SUL-1.0 license and free to install. However, it requires paid AI model subscriptions or API credits to function. Anonymous telemetry is enabled by default (single daily ping with hashed installation ID) but can be disabled via OMO_DISABLE_POSTHOG=1 or config setting.

What is ultrawork mode and how do I use it?+

Ultrawork mode is activated by typing 'ultrawork' or 'ulw' in your prompt. It triggers autonomous end-to-end execution where the main agent orchestrates specialists in parallel, doesn't stop until the task is complete, and handles planning, exploration, and execution without manual intervention. The Ultrawork Planner (/ulw-plan) provides interview-based strategic planning before code changes.

How do I install oh-my-openagent?+

Open the source repository on GitHub and follow its README. oh-my-openagent is a plugin — MCP Agents Market links you directly to the official repo.

Is oh-my-openagent free?+

oh-my-openagent is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.

Related searches