dmux Workflows
Orchestrate parallel AI agent sessions with dmux, a tmux pane manager for agents: research-plus-implement splits, multi-file features, test-and-fix loops, cross-harness workflows, review pipelines, git worktree isolation.
- What
- Orchestrate parallel AI agent sessions with dmux, a tmux pane manager for agents: research-plus-implement splits, multi-file features, test-and-fix loops, cross-harness workflows, review pipelines, git worktree isolation.
- Cost
- Free
- Needs
- Use "dmux Workflows" with your Muse.
- Install
- Copy the installer prompt below into your Muse — your agent does the rest.
Curated by Skill Harbor: a pattern guide for running multiple AI agent sessions in parallel with dmux, a tmux pane manager for agent harnesses (press n for a new pane with a prompt, m to merge output back). Five workflow patterns are spelled out with concrete pane assignments: research plus implement with findings merged into the builder's context; multi-file features split across schema, API, and UI panes with integration in the main pane; test plus fix loops with a watcher pane and a fixer pane; cross-harness work mixing Claude Code, Codex, and others by task fit; and parallel code review with security, performance, and coverage perspectives merged into one report. Best practices are firm: independent tasks only, clear file boundaries, strategic merging, git worktrees for conflict-prone work, and resource awareness since each pane is a full agent session, keep it to five or six panes. It also documents the ECC helper script for external tmux-pane orchestration with branch-backed worktrees, per-worker task and handoff files, and seed paths for dirty local files, plus a troubleshooting section for unresponsive panes, merge conflicts, high token usage, and missing tmux. By @affaan-m, listed here with credit to its creator. From the affaan-m/ECC repository (MIT). Honest caveats: dmux itself is a separate install from github.com/standardagents/dmux; every parallel pane burns API tokens, so small tasks are cheaper in a single session. Skill Harbor never reviews the code, review it yourself before use.
Version:
Install
Copy the install package below, then paste it into MuseThe install prompt below already includes the vetting steps: your agent follows the community checklist before installing anything with executable code. Want more?
Use "dmux Workflows" with your Muse. Prerequisites: dmux installed separately (review the package at https://github.com/standardagents/dmux first), tmux available, one task genuinely worth parallelizing. 1. Open the skill: https://github.com/affaan-m/ECC/blob/main/skills/dmux-workflows/SKILL.md and copy the full SKILL.md text. 2. Paste it into a chat with Muse and add: "Split this work into dmux panes: [describe the task]. Propose the pane plan first." 3. Review the pane plan, start dmux, and let Muse merge results back strategically. Tip: use git worktrees per pane when panes touch overlapping files; keep panes to five or six to control token spend. Safety: a skill is plain-text instructions; it runs nothing by itself. Review generated commands before running them.
Saved to your recent installs. Find it anytime on /connect.
Questions
How do I install a build?
Every product page includes a copy-paste install prompt. Paste it into your Muse and it sets the build up for you — no manual configuration.
Where does my money go?
Straight to the seller. Skill Harbor never processes payments: checkout happens on the seller’s own page, usually Stripe.
What does the ✓ next to a creator’s name mean?
It means we confirmed the identity of the person behind the listing. It says nothing about the code itself — always check a build before installing it.