探索
Gooseteam

Gooseteam

GooseTeam是一个AI代理协作平台,通过MCP协议实现代理间的任务分配与消息管理,支持自然语言和流程图两种协议表达方式,并提供扩展工具集。
2.5分
40
2025-04-28 23:14:13
概述
内容详情
替代品

What is GooseTeam MCP Server?

GooseTeam is a server that enables multiple AI agents to collaborate on tasks using the Model Context Protocol (MCP). It provides tools for agent registration, message management, and task coordination.

How to use GooseTeam?

1. Configure Goose CLI with the extension 2. Start the MCP server and inspector 3. Launch agents that follow the protocol 4. Monitor and interact through the inspector interface

Use Cases

Ideal for collaborative AI projects where multiple agents need to: - Divide and conquer complex tasks - Coordinate specialized roles - Maintain shared context and communication

Key Features

Agent ManagementRegister agents with unique IDs and colors, list active agents, and handle agent departures.
Message SystemSend, receive, and manage messages between agents with persistent storage.
Task CoordinationCreate, assign, and complete tasks with automatic role assignment (Project Coordinator).
Mermaid Protocol SupportBehavioral control through visual flowchart instructions using Mermaid markdown.

Pros and Cons

Advantages
Enables true multi-agent collaboration
Flexible protocol implementation (text or flowchart)
Visual monitoring through inspector interface
Language-agnostic MCP protocol
Limitations
Requires compatible LLM (currently only works well with certain models)
Setup requires multiple terminal sessions
Learning curve for protocol implementation

Getting Started

Install DependenciesClone the repository and install required packages.
Configure Goose CLIAdd the GooseTeam extension to your Goose configuration.
Start ServicesLaunch the server and inspector in separate terminals.
Launch AgentsStart agents that will connect to the server.

Usage Scenarios

Project CoordinationMultiple agents collaborate on a software development project
Research TeamAgents divide research topics and compile findings

Frequently Asked Questions

1
Why do my agents stop working after a few messages?This usually means your LLM isn't good at tool usage. Try different models - currently gemini-2.0-flash-001 and GPT-4o work best.
2
Can I use this with other agent platforms besides Goose?Yes, the protocol and MCP server aren't tightly coupled to Goose - only the npm scripts use Goose-specific commands.
3
How do I know if my agents are collaborating?Use the inspector interface to view messages and tasks in real-time.

Additional Resources

MCP SpecificationComplete Model Context Protocol documentation
Building a Team of AI Agents (Video)Inspiration talk by Aaron Goldsmith
Goose CLI DocumentationInstallation and usage guide for Goose
精选MCP服务推荐
Duckduckgo MCP Server
已认证
DuckDuckGo搜索MCP服务器,为Claude等LLM提供网页搜索和内容抓取服务
Python
206
4.3分
Firecrawl MCP Server
Firecrawl MCP Server是一个集成Firecrawl网页抓取能力的模型上下文协议服务器,提供丰富的网页抓取、搜索和内容提取功能。
TypeScript
2,951
5分
Figma Context MCP
Framelink Figma MCP Server是一个为AI编程工具(如Cursor)提供Figma设计数据访问的服务器,通过简化Figma API响应,帮助AI更准确地实现设计到代码的一键转换。
TypeScript
6,096
4.5分
Context7
Context7 MCP是一个为AI编程助手提供实时、版本特定文档和代码示例的服务,通过Model Context Protocol直接集成到提示中,解决LLM使用过时信息的问题。
TypeScript
4,851
4.7分
Exa Web Search
已认证
Exa MCP Server是一个为AI助手(如Claude)提供网络搜索功能的服务器,通过Exa AI搜索API实现实时、安全的网络信息获取。
TypeScript
1,425
5分
Edgeone Pages MCP Server
EdgeOne Pages MCP是一个通过MCP协议快速部署HTML内容到EdgeOne Pages并获取公开URL的服务
TypeScript
87
4.8分
Baidu Map
已认证
百度地图MCP Server是国内首个兼容MCP协议的地图服务,提供地理编码、路线规划等10个标准化API接口,支持Python和Typescript快速接入,赋能智能体实现地图相关功能。
Python
316
4.5分
Minimax MCP Server
MiniMax Model Context Protocol (MCP) 是一个官方服务器,支持与强大的文本转语音、视频/图像生成API交互,适用于多种客户端工具如Claude Desktop、Cursor等。
Python
359
4.8分
安装
复制以下命令到你的Client进行配置
注意:您的密钥属于敏感信息,请勿与任何人分享。