skills.rest
devseunggwan
devseunggwanCOMMUNITY
Explore all skills by this author →
GitHub

cmux-orchestrator

Coordinate and crash-isolate Claude Code workers across cmux workspaces.

Updated Jan 6, 2026
Search Tags:#automation#workflow#parallel#orchestrate#workspace-management#crash-recovery#cmux
One-click install
npx skills add https://github.com/devseunggwan/praxis --skill cmux-orchestrator
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: cmux-orchestrator
Source: https://github.com/devseunggwan/praxis/tree/main/skills/cmux-orchestrator
Command: npx skills add https://github.com/devseunggwan/praxis --skill cmux-orchestrator

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Dispatches and supervises multiple Claude Code workers in cmux workspaces, enabling crash-resilient, model-routed, file-based coordination so long-running tasks survive master failures.

Core Features & Use Cases

  • Crash-isolated workers with per-workspace state stored on disk, enabling restart and reattachment after failures.
  • Parallel task orchestration across multiple providers (Claude, Codex, Gemini) with provider routing and centralized monitoring.
  • Use cases include batch code reviews, large-scale test runs, migrations, and parallel automation pipelines that require reliable results collection.

Quick Start

Set up the orchestrator to run multiple cmux workers in parallel and monitor their progress until all complete.

Frequently Asked Questions about cmux-orchestrator

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run parallel tasks across multiple Claude Code workers without losing progress on crashes?▼

Parallel task orchestration dispatches Claude Code workers across cmux workspaces with file-based state, enabling crash-isolated execution and deterministic recovery after master failures.

What is file-based state coordination for multi-workspace automation pipelines?▼

File-based state coordination stores per-worker execution status on disk, allowing multiple cmux workers to operate independently and reattach to tasks after a crash.

Can I route parallel code reviews to different AI providers like Codex and Gemini?▼

Provider routing dispatches parallel tasks to Claude, Codex, or Gemini workers within the orchestration loop, allowing centralized monitoring and structured results collection across providers.

How do I set up crash-resilient batch test runs and migrations in parallel?▼

Set up the orchestrator to dispatch multiple cmux workers with per-workspace prompts, then monitor the poll-based supervision loop until all batch tasks complete and collect tiered outputs.

Does crash recovery for parallel workflows require external dependencies?▼

Crash recovery relies entirely on deterministic file-based state stored within cmux workspaces, requiring no external dependencies to restart and reattach workers after failures.

When should I use multi-workspace orchestration instead of a single worker?▼

Multi-workspace orchestration is ideal for large-scale batch code reviews, test runs, and migrations that require independent crash-isolated workers and reliable results collection.

Related Skills

Explore More Software Engineering Skills→
devseunggwandevseunggwan

cmux-recover-sessions

Scan persisted .jsonl files to recover crashed Claude Code sessions into cmux workspaces.

Community
Advanced
mejangomejango

sarif-parsing

Parse SARIF 2.1.0 files to extract and deduplicate findings.

Community
Intermediate
markus41markus41
1

llmintegration

Integrate Claude, GPT, Gemini, and Ollama APIs with a pluggable provider abstraction.

Community
Intermediate
marcusgollmarcusgoll

deep-execute

Processes bulk development task queues concurrently with Git conflict resolution and recovery support.

Community
Advanced