Rig

Community wiki for OpenCode and oh-my-openagent. Agent pipeline, skills ecosystem, playbooks, and setup guides — plus omo-kit, the CLI that generates your configs so you don't write JSON by hand.

What is this?

Two tools, one workflow:

• OpenCode: an open-source CLI and TUI that connects you to AI coding agents. It's the runtime: sessions, context, language models.
• oh-my-openagent (oMO): a configuration layer and skill ecosystem on top. Multi-agent pipeline (Prometheus → Atlas → Sisyphus), curated skills, project conventions.

Together they let you code with agents the way senior engineers work: plan first, delegate smart, verify everything.

Who is this for?

• Developers new to AI coding tools: you've heard about vibe coding and want a real workflow.
• OpenCode users looking to level up: you've run opencode but want the structured pipeline oMO provides.
• Teams standardizing on AI tooling: shared configs, project conventions, playbooks everyone can follow.

Quickstart

1. Install OpenCode: get the CLI running on your machine.
2. Install oMO: add the pipeline, skills, and conventions.
3. Your First Session: open a project, ask a question, see the pipeline.

Explore

• The oMO Pipeline: how Prometheus, Atlas, and Sisyphus work together.
• Skills Ecosystem: built-in skills, categories, Matt Pocock, MCP, DIY.
• Playbooks: real workflows for real tasks.
• Reference: deep config docs for oh-my-openagent.json, model selection, and more.

Start here

Every project oMO knows about uses two files in the repo root:

• CLAUDE.md: project instructions for Claude Code sessions.
• AGENTS.md: project instructions for Codex / OpenCode agents.

These files tell agents about your stack, conventions, and rules. oMO reads them automatically: you don't need to paste context every session.

Get started →