dmux-workflows

Coordinate parallel AI agent sessions across development harnesses using tmux panes and git worktrees.

Updated May 9, 2026
One-click install
npx skills add https://github.com/kk20300113-png/my-claude-skills --skill dmux-workflows-kk20300113-png
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: dmux-workflows
Source: https://github.com/kk20300113-png/my-claude-skills/tree/main/dmux-workflows
Command: npx skills add https://github.com/kk20300113-png/my-claude-skills --skill dmux-workflows-kk20300113-png

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill eliminates the manual overhead of coordinating multiple parallel AI agent sessions across different development tools, reducing context switching and speeding up complex development workflows that would take much longer to complete sequentially.

Core Features & Use Cases

  • Cross-Harness Agent Orchestration: Run parallel agent sessions across Claude Code, Codex, OpenCode, Cline, Gemini, and Qwen in isolated tmux panes.
  • Pre-Built Workflow Patterns: Includes ready-to-use patterns for common multi-agent tasks like research + implementation, multi-file feature development, test-fix loops, and parallel code review.
  • Conflict Avoidance: Integrates with git worktrees to isolate file changes between parallel agents, eliminating merge conflicts.
  • Use Case: For a complex e-commerce checkout feature, you can run parallel agents to implement the payment processing logic, build the UI components, and write end-to-end tests simultaneously, then merge all results into your main codebase.

Quick Start

Use the dmux-workflows skill to split your new user authentication feature into parallel agent sessions for security research, middleware implementation, and unit test writing, then merge the completed work into your main branch.

Frequently Asked Questions about dmux-workflows

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

FAQPage Schema
How do I run parallel AI agent workflows without file conflicts?

Run parallel AI agent workflows without file conflicts by orchestrating sessions across isolated tmux panes and using git worktrees to separate file changes between agents, eliminating merge conflicts.

Can I use tmux and git worktrees to orchestrate multiple AI coding agents?

Yes, you can use tmux and git worktrees to orchestrate multiple AI coding agents by running isolated sessions in separate panes and managing branch isolation to prevent file overwrites during parallel development.

What is the best way to coordinate parallel coding sessions across different AI harnesses?

The best way to coordinate parallel coding sessions across different AI harnesses is using pre-defined workflow patterns that manage divide-and-conquer tasks like research, implementation, and test-fix loops across tools like Claude Code and Codex.

How do I split a complex software feature into parallel AI agent tasks?

Split a complex software feature into parallel AI agent tasks by applying pre-built workflow patterns that assign isolated git worktrees and tmux panes to specific sub-tasks like UI development, backend logic, and testing simultaneously.

Does parallel multi-agent orchestration work with different AI development tools?

Parallel multi-agent orchestration works with different AI development tools by running cross-harness sessions across Claude Code, Codex, OpenCode, Cline, Gemini, and Qwen within isolated tmux panes for concurrent execution.

When should I avoid using parallel agent workflows for software development?

Avoid using parallel agent workflows for software development when tasks are highly interdependent and require sequential logic, as divide-and-conquer parallelism works best for isolated sub-tasks like independent code reviews or separate feature modules.