LaunchKit · 2026
Back to Skills

Manage multiple workplaces (project directories) with multi-agent orchestration, isolated memory, and inter-agent.

0
388 downloads
by @dickwu

Setup & Installation

openclaw skills install @dickwu/workplace

Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:

npx clawhub install workplace

What This Skill Does

Manages multiple project directories as named workplaces with isolated memory, per-workspace agents, and session tracking. Each workplace stores its own agent definitions, deployment configs, and chat history under a .workplace/ directory. Supports switching between projects, linking related workplaces, and syncing context to external IDEs.

Keeps agent memory, skills, and session state scoped per project so switching between codebases does not bleed context from one workspace into another.

When to use it

  • Switching between client projects without losing agent context
  • Running a reviewer agent in one codebase while a coder agent works in another
  • Exporting a workspace config to share project setup with a teammate
  • Syncing active project context to Cursor or Claude Code before a coding session
  • Resuming a saved chat session after returning to a project the next day

Example Workflow

Here's how your AI assistant might use this skill in practice.

INPUT

User asks: switch to my log-stream project and show me what agents are available

AGENT
  1. 1Read current.json to identify the active workplace
  2. 2Run workplace switch log-stream to update current.json and registry.json
  3. 3Load .workplace/config.json from the log-stream directory
  4. 4Check sessions.json for existing sessions and present Continue or New chat options
  5. 5Run workplace agents to list agent definitions from .workplace/agents/
OUTPUT

Confirmation message showing the switched workplace name, path, linked workplaces, and a list of available agents with start/stop buttons