Overview
start-work is an open-source workflow skill for Claude Code and compatible agents, published by code-yeongyu. Its author describes it as: “Execute a Prometheus work plan with Boulder state, evidence ledger updates, worktree discipline, parallel subagents, and Stop-hook continuation. Use after planning when the user says start work, execute plan, continue…”. The project has 68k stars on GitHub and ships custom license terms, so review them before reuse. Add it to your setup with `git clone https://github.com/code-yeongyu/oh-my-openagent ~/.claude/skills/start-work`.
What start-work does
**YOU DO NOT WRITE CODE. YOU DO NOT EDIT PRODUCT FILES. YOU DO NOT RUN QA YOURSELF. EVERY unit of implementation, test, QA, and review work MUST be delegated to a spawned subagent. NO EXCEPTIONS.** Your hands touch only plan selection, `.omo/` state (Boulder, ledger, plan checkboxes), decomposition, dispatch, verdicts, and evidence records. About to edit a product file or run an implementation command yourself? **STOP. SPAWN A WORKER INSTEAD.** Orchestrate at **MAXIMUM PARALLELISM**: every independent unit runs concurrently; only named dependencies serialize.
What’s inside
The SKILL.md for start-work is organised into these sections:
- ABSOLUTE RULE: YOU ARE AN ORCHESTRATOR — NEVER THE IMPLEMENTER
- Codex Harness Tool Compatibility
- Codex tier mapping for the delegation router
- Codex Subagent Reliability
- Usage
- Goal and todo discipline (MANDATORY)
- Phase 1: Select the plan
- No-plan bootstrap
- Phase 2: Create or update Boulder state
- Parallel delivery lanes (teams and worktrees)
- Phase 3: Execute the next checkbox
- Monitor every dispatched subagent to its completion condition
Installation
Add start-work to your agent with:
git clone https://github.com/code-yeongyu/oh-my-openagent ~/.claude/skills/start-workReview a skill’s source before installing it — skills can instruct your agent to run commands. This command comes from the skill’s public repository, which is the source of truth for setup steps. New to skills? Read how to install Claude Code skills.
Supported agents
- Codex — declared in the repository topics
- Cursor — declared in the repository topics
- OpenCode — declared in the repository topics
License
code-yeongyu has not declared a standard open-source license for this repository (Other). Confirm the terms with the author before reusing it.
Strengths
- Ships in code-yeongyu/oh-my-openagent, an established project with 68,307 GitHub stars
- Works with Codex, Cursor, OpenCode
- Actively maintained — repository updated in the last 90 days
- Its repository gained 1,416 GitHub stars over the last 24 days
Things to check
- No license declared — confirm terms with the author before reuse
Author
Topics
Frequently asked questions
- What does start-work do?
- Execute a Prometheus work plan with Boulder state, evidence ledger updates, worktree discipline, parallel subagents, and Stop-hook continuation. Use after planning when the user says start work, execute plan, continue plan, resume plan, or asks to run a .omo/plans plan.
- How do I install start-work?
- Run git clone https://github.com/code-yeongyu/oh-my-openagent ~/.claude/skills/start-work in your agent, then reload your skills. Review the source at https://github.com/code-yeongyu/oh-my-openagent before installing.
- Is start-work free to use?
- start-work is publicly available on GitHub, but code-yeongyu has not declared a standard open-source license (Other). Confirm the terms with the author before reusing it.
- Which agents does start-work work with?
- start-work lists Codex, Cursor, and OpenCode — named in its SKILL.md, its plugin marketplace or its repository topics. Skills use the portable SKILL.md format, so other agents that read that format may load it too.
- Where does start-work come from?
- start-work ships inside code-yeongyu/oh-my-openagent, a repository that contains 49 catalogued skills in total. The repository's 68,307 GitHub stars apply to that whole collection, not to this skill on its own.