Choose the right isolation strategy when spinning parallel cmux agents — shared filesystem, file ownership, or git worktrees.
Write effective standalone prompts for cmux subagents. Subagents have no memory of the current conversation — every prompt must be a self-contained spec.
Spin up parallel Claude subagents in cmux panes, delegate tasks, monitor, and close when done.
Re-attach Claude Code sessions to restored cmux panes after a reboot. Records cwd → session_id on Claude start and replays `claude --resume <id>` into each restored pane.
Personal Claude Code plugin repository for mjmendo. Add new skills here and they'll be available across all machines after a quick install.
1. Register this repo as a marketplace
Add the following to ~/.claude/settings.json (create the file if it doesn't exist):
{
"extraKnownMarketplaces": {
"mjmendo": {
"source": {
"source": "github",
"repo": "mjmendo/claude-plugins"
}
}
}
}
2. Install the plugin
Open Claude Code and run:
/plugin install mjmendo-plugins@mjmendo
That's it. All skills are now available.
Skills are invoked with a slash command. After installation, the cmux skills are available as:
/cmux-agents
/cmux-agents-isolation
/cmux-agents-prompts
/cmux-claude-resume
Claude also invokes skills automatically when relevant — e.g., if you ask to spin up parallel agents, it will use /cmux-agents without you having to call it explicitly.
1. Create the skill directory and file
skills/
my-skill/
SKILL.md ← required
reference.md ← optional supporting docs
2. Write SKILL.md
The frontmatter controls how Claude uses the skill:
---
description: One-line description — Claude uses this to decide when to invoke the skill.
allowed-tools: Read, Bash
---
Your instructions here. Use $ARGUMENTS to reference what the user passed.
3. Push to this repo
git add skills/my-skill/
git commit -m "Add my-skill"
git push
4. Update on your machines
/plugin update mjmendo-plugins@mjmendo
.claude-plugin/
plugin.json ← plugin manifest (name, version, paths)
skills/
cmux-agents/
SKILL.md
cmux-agents-isolation/
SKILL.md
cmux-agents-prompts/
SKILL.md
cmux-claude-resume/
SKILL.md
install.sh
uninstall.sh
bin/
LaunchAgents/
README.md
| Skill | Invoke | Description |
|---|---|---|
cmux-agents | /cmux-agents | Orchestrate parallel Claude subagents in cmux panes |
cmux-agents-isolation | /cmux-agents-isolation | Pick the right isolation strategy (shared fs, file ownership, worktrees) |
cmux-agents-prompts | /cmux-agents-prompts | Write self-contained prompts for subagents |
cmux-claude-resume | /cmux-claude-resume | Re-attach Claude sessions to cmux panes after a reboot (one-time install) |
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
npx claudepluginhub mjmendo/claude-plugins --plugin mjmendo-pluginsUnified capability management center for Skills, Agents, and Commands.
Skill and agent authoring tools: generate, audit, and improve Claude Code skills, commands, and agents
Automatically detects and loads AGENTS.md files to provide agent-specific instructions alongside CLAUDE.md. Enables specialized agent behaviors without manual intervention.
Manage Claude Code settings and MCP server configurations with best practices
Dynamic orchestrator plugin — 200+ agents, 200+ skills, hooks, and MCP servers. Upstream sources linked via git submodules (agency-agents, everything-claude-code, oh-my-claudecode, gstack, superpowers)
Analyze codebases and recommend tailored Claude Code automations such as hooks, skills, MCP servers, and subagents.