Claude Code tmux Runner#

Overview#

Run multiple Claude Code tasks in parallel using tmux sessions. Non-blocking, background execution with status monitoring.

Usage#

Start a Task#

bash

~/.openclaw/scripts/claude-tmux.sh start "Your task description" [task-name]

Examples:

bash

~/.openclaw/scripts/claude-tmux.sh start "Write a Python auth module" auth
~/.openclaw/scripts/claude-tmux.sh start "Build a React component" ui
~/.openclaw/scripts/claude-tmux.sh start "Create a Go microservice" api

List All Tasks#

bash

~/.openclaw/scripts/claude-tmux.sh list

Check Task Status#

bash

~/.openclaw/scripts/claude-tmux.sh status <task-id>

View Generated Files#

bash

~/.openclaw/scripts/claude-tmux.sh files

Stop Tasks#

bash

~/.openclaw/scripts/claude-tmux.sh stop <task-id>
~/.openclaw/scripts/claude-tmux.sh stop-all

Workflow#

1. User says "use Claude Code xxx" → Optimize prompt → Show for approval
2. User confirms → Start background task → Return immediately
3. User can continue chatting → Non-blocking
4. User asks "how is Claude doing" → Check status → Return progress

Industrial Code Standards#

Generated code must include:

1. Complete error handling (try-catch, boundary checks)
2. Type annotations (TypeScript, Python type hints)
3. Documentation (JSDoc, docstring)
4. Unit tests (pytest, jest)
5. Performance considerations (time/space complexity)
6. Security best practices (input validation)
7. Maintainability (clear naming, modular)

File Locations#

• Script: ~/.openclaw/scripts/claude-tmux.sh
• Logs: ~/.openclaw/logs/claude/
• Files: ~/.openclaw/workspace/
• State: ~/.openclaw/state/claude/