A high-performance background task manager that enables parallel execution of Claude Code agents using tmux.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install claude-tmux-runner
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install claude-tmux-runner using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Claude Code tmux Runner?
The Claude Code tmux Runner is a specialized tool designed to handle concurrent execution of AI-driven coding tasks. By utilizing tmux as a backend, it allows developers to spin up multiple background sessions for Claude Code, effectively turning a linear workflow into a parallel powerhouse. This skill is essential for users of Openclaw Skills who need to offload long-running tasks like module generation, large-scale refactoring, or test suite creation without locking their primary interface.
Built for the modern developer workflow, it ensures that background jobs remain isolated, observable, and manageable. Whether you are building a React frontend, a Python auth module, and a Go microservice simultaneously, this skill manages the lifecycle of each agent process, providing a non-blocking environment that maximizes productivity.
Claude Code tmux Runner Use Cases
- Running multiple Claude Code tasks simultaneously without blocking the terminal session.
- Offloading time-intensive code generation or refactoring tasks to background processes.
- Monitoring the real-time progress of several autonomous coding agents across different project modules.
- Managing a suite of background AI jobs through a centralized CLI within the Openclaw Skills ecosystem.
How Claude Code tmux Runner Works
- The process begins when a user requests a task using Claude Code which the system identifies as a candidate for background execution.
- The skill optimizes the prompt and presents it to the user for final approval.
- Upon confirmation, the skill triggers the
claude-tmux.shscript to spawn a new, named tmux session. - The agent executes the task in the background, allowing the user to continue chatting or initiate other tasks immediately.
- Users can query the status of specific tasks or list all active jobs to monitor progress.
- Once complete, generated files are moved to the designated workspace and logs are archived for review.
Claude Code tmux Runner Setup
To use this skill, ensure tmux is installed on your system. The core script is automatically integrated into the environment.
# Start a background task with a specific name
~/.openclaw/scripts/claude-tmux.sh start "Write a Python auth module" auth
# Check the status of your active sessions
~/.openclaw/scripts/claude-tmux.sh list
Claude Code tmux Runner Data Schema & Taxonomy
The skill manages its operational data across a structured directory hierarchy to ensure persistence and traceability:
| Location | Description |
|---|---|
~/.openclaw/scripts/claude-tmux.sh |
The main execution script for session management. |
~/.openclaw/logs/claude/ |
Stores detailed execution logs (stdout/stderr) for every task. |
~/.openclaw/workspace/ |
The target directory where all code generated by the agent is stored. |
~/.openclaw/state/claude/ |
Tracks the metadata, PIDs, and session names of background tasks. |
Claude Code tmux Runner Advanced Features
- Isolated parallel execution via independent tmux sessions for every task.
- Centralized task management including
stop-alland detailed status retrieval commands. - Non-blocking architecture that allows for continuous interaction with other Openclaw Skills while agents work.
- Enforced industrial code standards, ensuring all background-generated code includes error handling, type annotations, and unit tests.
- Automatic prompt optimization before session hand-off to ensure high-quality agent outputs.
SKILL.md
Loading
Related Openclaw Skills
Cache Cleanup
An automated system maintenance utility designed to purge expired logs, temporary files, and cached data to maintain peak system performance.
Auto Workflow
Auto Workflow is a proactive tool designed to identify and transform repetitive manual tasks into efficient, automated execution scripts.
Auto Monitor
A proactive system monitoring tool that detects server health issues and reports them automatically without manual commands.
Raycast Extensions Skill
A technical framework for building and maintaining Raycast extensions using React, TypeScript, and the Raycast API.
doc-handler
A comprehensive Python-based utility for programmatically reading and editing Word, PDF, and Excel documents via a CLI.
Jina Web Fetcher
A high-performance utility that leverages Jina AI to fetch and convert any webpage or Google search result into clean, LLM-friendly markdown content.
Featured*
