cowork-sessions

Orchestrate multiple plugin agents across sequential, parallel, and fan-out/fan-in coordination models.

18|2|Updated Dec 12, 2025
One-click install
npx skills add https://github.com/markus41/claude --skill cowork-sessions
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: cowork-sessions
Source: https://github.com/markus41/claude/tree/main/plugins/cowork-marketplace/skills/cowork-sessions
Command: npx skills add https://github.com/markus41/claude --skill cowork-sessions

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill provides the knowledge and framework for launching, managing, and monitoring complex cowork sessions that coordinate multiple AI agents working in parallel to achieve a larger goal.

Core Features & Use Cases

  • Session Lifecycle Management: Understand and control the stages of a cowork session from initialization to completion.
  • Agent Coordination: Learn patterns for sequential, parallel, and fan-out/fan-in execution of agents.
  • Resource Monitoring: Track tokens, cost, and duration for optimized resource utilization.
  • Use Case: Launch a "market research" cowork session that simultaneously tasks agents to analyze competitor websites, summarize recent news, and gauge social media sentiment, then merges their findings into a comprehensive report.

Quick Start

Launch a cowork session to analyze competitor websites and summarize recent news.

Frequently Asked Questions about cowork-sessions

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I orchestrate parallel AI agents for complex tasks?

To orchestrate parallel AI agents, you manage a cowork session lifecycle from initialization through completion, coordinating execution across sequential, parallel, and fan-out/fan-in patterns to achieve complex workflow goals.

What is the fan-out/fan-in pattern for agent coordination?

The fan-out/fan-in pattern dispatches multiple agents to execute parallel subtasks, then merges their independent findings into a unified result, enabling efficient execution of complex multi-agent workflows.

Can I track token cost and duration during multi-agent workflow automation?

Yes, you can track resource utilization including tokens, cost, and duration during multi-agent workflow automation, enabling precise monitoring and optimization of parallel session execution metrics.

How do I initialize a cowork session for multiple plugin agents?

You initialize a cowork session by defining the orchestration plan for multiple plugin agents, establishing sequential, parallel, or fan-out/fan-in execution coordination models before launching complex task workflows.

Does parallel agent orchestration handle error handling across different execution models?

Yes, parallel agent orchestration includes error handling across sequential, parallel, and fan-out/fan-in coordination models, ensuring robust session completion and workflow stability even when individual plugin agents encounter failures.

What is the best way to coordinate multiple AI agents for market research?

The best way to coordinate multiple AI agents for market research is using a fan-out/fan-in cowork session, simultaneously tasking agents to analyze competitors, summarize news, and gauge sentiment before merging findings into a comprehensive report.