Yeachan-Heo/oh-my-claudecode Instruction Manual

What Yeachan-Heo/oh-my-claudecode is—in plain language

Teams-first multi-agent orchestration layer for Claude Code with guided setup, staged autopilot, native teams, provider advisers, planning, QA, memory, diagnostics, and many focused skills.

Mental model: This repository extends a host application. The host loads its commands, skills, hooks, or agents; most day-to-day use happens inside that host rather than by launching a separate app.

InterfacesClaude slash commands, natural-language triggers, omc CLI

Primary languageTypeScript, JavaScript, Shell, Markdown

Typical useRun setup, use deep-interview or plan for ambiguous work, and use team/autopilot only after reviewing scope and permissions

Your first 15 minutes

Complete the linked Installation Guide and its verification step.
Create a disposable test folder or use non-sensitive sample data.
Run the small example below and observe what files, ports, or prompts appear.
Read the result before approving writes, network calls, account access, or costs.
Only then repeat the workflow with a real project.

Starter workflow

# Inside Claude Code, run:
/deep-interview "Plan a tiny, disposable command-line notes app"

Open Installation Guide Open upstream README

Complete indexed functionality map

Each card below corresponds to a component identified in the repository research. Open a component record for its path, purpose, capabilities, dependencies, risks, and relationships.

ai-slop-cleaner

skill

Remove low-information AI mannerisms and noisy boilerplate while preserving the intended technical result.

skills/ai-slop-cleaner/SKILL.md

Open component record →

ask

skill

Send a bounded question to a configured external provider and bring the response back into the active workflow.

skills/ask/SKILL.md

Open component record →

autopilot

skill

Run a staged plan-build-verify workflow from a high-level goal with checkpoints and recoverable state.

skills/autopilot/SKILL.md

Open component record →

autoresearch

skill

Execute an iterative, evidence-recorded research loop for measurable technical experiments.

skills/autoresearch/SKILL.md

Open component record →

cancel

skill

Stop an active OMC mode or workflow and clean up its tracked state safely.

skills/cancel/SKILL.md

Open component record →

ccg

skill

Ask multiple coding providers for complementary analysis and have Claude synthesize their findings.

skills/ccg/SKILL.md

Open component record →

configure-notifications

skill

Configure desktop or external notifications for long-running OMC activity.

skills/configure-notifications/SKILL.md

Open component record →

debug

skill

Drive a reproduction-first debugging workflow that traces symptoms to root cause and verifies the fix.

skills/debug/SKILL.md

Open component record →

deep-dive

skill

Perform a deliberate, high-context investigation of a codebase, architecture, or difficult technical question.

skills/deep-dive/SKILL.md

Open component record →

deep-interview

skill

Clarify an ambiguous project through structured questions before planning or implementation begins.

skills/deep-interview/SKILL.md

Open component record →

deepinit

skill

Create a deeper project initialization and context map for subsequent agent work.

skills/deepinit/SKILL.md

Open component record →

external-context

skill

Gather explicitly requested outside context and package it for the active Claude Code session.

skills/external-context/SKILL.md

Open component record →

hud

skill

Display compact workflow, context, and agent-status information during a session.

skills/hud/SKILL.md

Open component record →

learner

skill

Capture reusable lessons from completed work and convert them into future guidance.

skills/learner/SKILL.md

Open component record →

local-build-reminder

skill

Keep source-build users aware when generated or compiled artifacts must be refreshed.

skills/local-build-reminder/SKILL.md

Open component record →

mcp-setup

skill

Configure supported MCP integrations and verify that Claude Code can reach them.

skills/mcp-setup/SKILL.md

Open component record →

merge-readiness

skill

Assess tests, review findings, conflicts, and delivery evidence before a branch or pull request is merged.

skills/merge-readiness/SKILL.md

Open component record →

omc-doctor

skill

Diagnose installation, configuration, runtime, hook, and dependency problems in OMC.

skills/omc-doctor/SKILL.md

Open component record →

omc-reference

skill

Open the consolidated OMC command, mode, configuration, and compatibility reference.

skills/omc-reference/SKILL.md

Open component record →

omc-setup

skill

Install and configure OMC's project and user-level files through a guided setup flow.

skills/omc-setup/SKILL.md

Open component record →

omc-teams

skill

Compatibility entry point that directs work to the current team orchestration surface.

skills/omc-teams/SKILL.md

Open component record →

plan

skill

Turn a request into an ordered implementation plan with explicit scope and verification steps.

skills/plan/SKILL.md

Open component record →

project-session-manager

skill

Create and manage project sessions, worktrees, templates, and provider-specific execution contexts.

skills/project-session-manager/SKILL.md

Open component record →

ralph

skill

Persist on an implementation goal through bounded retries and verification until its completion condition is met.

skills/ralph/SKILL.md

Open component record →

ralplan

skill

Iterate on a plan with critique and refinement before expensive execution begins.

skills/ralplan/SKILL.md

Open component record →

release

skill

Coordinate versioning, validation, changelog, packaging, and release-readiness work.

skills/release/SKILL.md

Open component record →

remember

skill

Save an explicit fact or decision to OMC's durable project memory.

skills/remember/SKILL.md

Open component record →

sciomc

skill

Apply a scientific hypothesis-experiment-measurement loop to technical investigation.

skills/sciomc/SKILL.md

Open component record →

self-improve

skill

Evaluate agent behavior against goals and propose measured configuration or workflow improvements.

skills/self-improve/SKILL.md

Open component record →

setup

skill

Provide the short in-session entry point to OMC installation and configuration.

skills/setup/SKILL.md

Open component record →

skill

Inspect, select, and use an installed OMC skill deliberately.

skills/skill/SKILL.md

Open component record →

skillify

skill

Convert a repeatable workflow or source document into a structured reusable skill.

skills/skillify/SKILL.md

Open component record →

team

skill

Coordinate Claude Code's native multi-agent team through planning, execution, verification, and repair stages.

skills/team/SKILL.md

Open component record →

trace

skill

Trace agent decisions, hooks, state transitions, or execution paths for diagnosis.

skills/trace/SKILL.md

Open component record →

ultragoal

skill

Manage a large goal with durable state, decomposition, progress checks, and completion criteria.

skills/ultragoal/SKILL.md

Open component record →

ultraqa

skill

Run a deeper quality-assurance pass across behavior, tests, edge cases, and deliverables.

skills/ultraqa/SKILL.md

Open component record →

ultrawork

skill

Use higher-effort parallel execution for a broad, well-scoped implementation workload.

skills/ultrawork/SKILL.md

Open component record →

verify

skill

Check claims and deliverables against concrete tests, commands, artifacts, or acceptance criteria.

skills/verify/SKILL.md

Open component record →

visual-verdict

skill

Evaluate visual output against explicit design and usability criteria.

skills/visual-verdict/SKILL.md

Open component record →

wiki

skill

Build and query durable project documentation from repository evidence.

skills/wiki/SKILL.md

Open component record →

writer-memory

skill

Maintain writing-specific preferences, facts, terminology, and continuity across sessions.

skills/writer-memory/SKILL.md

Open component record →

Upstream feature-by-feature guide

These capabilities are derived from the current first-party README and supporting documentation. Names follow upstream terminology so you can search the source documentation precisely.

Upstream capability

Core Maintainers

Upstream treats this as a distinct part of Yeachan-Heo/oh-my-claudecode. In plain language, use this area when your task concerns core maintainers. Begin with the documented default, test it on disposable input, and open the source section for its current options and limitations.

Upstream capability

Ambassadors

Upstream treats this as a distinct part of Yeachan-Heo/oh-my-claudecode. In plain language, use this area when your task concerns ambassadors. Begin with the documented default, test it on disposable input, and open the source section for its current options and limitations.

Upstream capability

Document Specialists

Upstream treats this as a distinct part of Yeachan-Heo/oh-my-claudecode. In plain language, use this area when your task concerns document specialists. Begin with the documented default, test it on disposable input, and open the source section for its current options and limitations.

Upstream capability

Top Collaborators

Upstream treats this as a distinct part of Yeachan-Heo/oh-my-claudecode. In plain language, use this area when your task concerns top collaborators. Begin with the documented default, test it on disposable input, and open the source section for its current options and limitations.

Upstream capability

CLI Commands vs In-Session Skills

Upstream treats this as a distinct part of Yeachan-Heo/oh-my-claudecode. In plain language, use this area when your task concerns cli commands vs in-session skills. Begin with the documented default, test it on disposable input, and open the source section for its current options and limitations.

Upstream capability

VS Code, Agent SDK, and automation scope

Upstream treats this as a distinct part of Yeachan-Heo/oh-my-claudecode. In plain language, use this area when your task concerns vs code, agent sdk, and automation scope. Begin with the documented default, test it on disposable input, and open the source section for its current options and limitations.

Upstream capability

Not Sure Where to Start?

Upstream treats this as a distinct part of Yeachan-Heo/oh-my-claudecode. In plain language, use this area when your task concerns not sure where to start?. Begin with the documented default, test it on disposable input, and open the source section for its current options and limitations.

Upstream capability

Team Mode (Recommended)

Upstream treats this as a distinct part of Yeachan-Heo/oh-my-claudecode. In plain language, use this area when your task concerns team mode (recommended). Begin with the documented default, test it on disposable input, and open the source section for its current options and limitations.

Upstream capability

tmux CLI Workers — Codex, Gemini & Antigravity (v4.4.0+)

Upstream treats this as a distinct part of Yeachan-Heo/oh-my-claudecode. In plain language, use this area when your task concerns tmux cli workers — codex, gemini & antigravity (v4.4.0+). Begin with the documented default, test it on disposable input, and open the source section for its current options and limitations.

Upstream capability

Updating

Upstream treats this as a distinct part of Yeachan-Heo/oh-my-claudecode. In plain language, use this area when your task concerns updating. Begin with the documented default, test it on disposable input, and open the source section for its current options and limitations.

Upstream capability

Why oh-my-claudecode?

Upstream treats this as a distinct part of Yeachan-Heo/oh-my-claudecode. In plain language, use this area when your task concerns why oh-my-claudecode?. Begin with the documented default, test it on disposable input, and open the source section for its current options and limitations.

Upstream capability

Orchestration Modes

Upstream treats this as a distinct part of Yeachan-Heo/oh-my-claudecode. In plain language, use this area when your task concerns orchestration modes. Begin with the documented default, test it on disposable input, and open the source section for its current options and limitations.

How to use the main workflows

Choose the smallest relevant function

Start with one capability card instead of asking the repository to do everything at once. This makes permissions, inputs, and output easier to understand.

Prepare a disposable input

Use a sample URL, copied repository, test document, or sandbox account. Keep production credentials and irreplaceable files out of the first run.

Run, observe, and stop

Watch Terminal output or the host application's activity view. If the behavior differs from the README, stop with Control + C or cancel inside the host before retrying.

Inspect the result

Check generated files, diffs, API responses, logs, or previews. Never assume “command finished” means the result is correct.

Save a repeatable recipe

Record the working command and non-secret configuration in your project's README. Store secrets in the documented environment file or password manager.

Copy-paste recipes from upstream documentation

These examples are selected from the repository's first-party documentation. Replace obvious placeholders, keep quotation marks intact, and run them only in the context indicated by the surrounding explanation.

Not Sure Where to Start?

/deep-interview "I want to build a task management app"

→ When prompted, type "openclaw" → choose "OpenClaw Gateway"

{
  "enabled": true,
  "gateways": {
    "my-gateway": {
      "url": "https://your-gateway.example.com/wake",
      "headers": { "Authorization": "Bearer YOUR_TOKEN" },
      "method": "POST",
      "timeout": 10000
    }
  },
  "hooks": {
    "session-start": {
      "gateway": "my-gateway",
      "instruction": "Session started for {{projectName}}",
      "enabled": true
    },
    "stop": {
      "gateway": "my-gateway",
      "instruction": "Session stopping for {{projectName}}",
      "enabled": true
    }
  }
}

Configuration, accounts, and files

Configuration files

.claude/omc.jsonc and ~/.config/claude-omc/config.jsonc

A configuration file changes behavior without changing source code. Make one change at a time and keep a backup before editing JSON, TOML, YAML, or environment files.

Important directories

skills/, scripts/, docs/

Paths identify where the relevant implementation or generated files live. Paths beginning with ~ are inside your home folder.

Credentials

Claude Code authentication and optional credentials for Codex, Gemini, Antigravity, Grok, or Cursor advisers/workers

Prefer temporary, least-privileged credentials. Never commit .env, tokens, cookies, private keys, or session exports.

External services

Selected model/provider CLIs and notification integrations

Check pricing, data retention, rate limits, and account permissions before enabling optional integrations.

Safe operating habits

Use test data first and keep a current backup or Git commit.
Read commands before pasting. A README is useful evidence, not a substitute for judgment.
Review agent-generated file changes with git diff before committing.
Keep local services bound to 127.0.0.1 unless you intentionally secure and expose them.
Do not give plugins or MCP servers broader filesystem, browser, GitHub, or cloud access than their current task requires.
Confirm API costs and model names before running large batches.
Stop and investigate repeated authentication failures instead of pasting a token into multiple places.

Reference and source trail

The manual reflects first-party files checked on August 18, 2026. It explains the indexed repository rather than promising that every optional third-party integration is available or safe.