mcp-agent-forge

Agent Forge - 智能体锻造工具 (AI Agent Forge Tool)

English | 中文

中文版

Agent Forge 是一个智能体创建和管理平台，能够创建和管理具有特定性格特征的智能体，并模拟它们对问题的回答。通过Agent Forge MCP，你可以快速构建起一个类似于CO-STORM的多智能体协作研究项目。

功能特点

智能体锻造：创建具有特定性格特征的智能体
思维模拟：模拟智能体回答问题
完整管理：支持智能体的查询、列表、删除等操作
多轮对话：支持深度的多轮对话交互
自然语言处理：基于 DeepSeek API 的高级语言理解能力

系统要求

Go 1.24.1 或更高版本
DeepSeek API 密钥

安装

git clone https://github.com/HundunOnline/mcp-agent-forge.git
cd mcp-agent-forge && make build

MCP 配置

{
  "mcpServers": {
    "mcp-agent-forge": {
      "command": "/path/to/mcp-agent-forge",
      "env": {
        "DEEPSEEK_API_KEY": "xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxx",
      }
    }
  }
}

配置说明

Configuration

Environment Variables

变量名	描述	默认值	是否必需
`DEEPSEEK_API_KEY`	DeepSeek API 密钥	-	是
`LOG_LEVEL`	日志级别 (debug, info, warn, error)	info	否
`LOG_PATH`	日志文件路径	./logs	否
`CONFIG_PATH`	配置文件路径	./config/config.yaml	否
`PORT`	服务端口号	8080	否
`DEBUG`	调试模式开关	false	否

使用方法

expert_personality_generation: 创建新的智能体
agent_answer: 模拟智能体回答问题
get_agent: 获取智能体信息
list_agents: 列出所有智能体
delete_agent: 删除智能体

示例

基本用法

// 创建智能体
{
    "name": "expert_personality_generation",
    "arguments": {
        "agent_name": "马斯克思维模型",
        "core_traits": "系统思维,第一性原理,工程思维,风险管理,创新思维"
    }
}

// 智能体回答
{
    "name": "agent_answer",
    "arguments": {
        "agent_id": "your_agent_id",
        "context": "如何看待特斯拉的发展策略？",
        "planned_rounds": 3,
        "current_round": 1,
        "need_more_rounds": false
    }
}

实际应用案例

我们在 Claude AI 中创建了一个示例应用，展示了如何使用 Agent Forge 创建和管理专家智能体：

Claude AI Demo

这个示例展示了：

如何创建具有特定专业背景的智能体
如何进行多轮对话交互
如何利用智能体的专业知识解决问题
如何管理和调整智能体的行为

贡献指南

欢迎提交 Pull Request 或创建 Issue 来帮助改进这个项目。我们特别欢迎以下方面的贡献：

新的智能体模型和特征
性能优化
文档改进
Bug 修复
新功能建议

许可证

本项目采用 MIT 许可证。详见 LICENSE 文件。

English Version

Agent Forge is a platform for creating and managing AI agents with specific personality traits and simulating their responses to questions. Through agent forge mcp, you can quickly build a multi-agent collaboration research project similar to CO-STORM.

Features

Agent Forging: Create agents with specific personality traits
Thought Simulation: Simulate agent responses to questions
Complete Management: Support for agent querying, listing, deletion, and other operations
Multi-round Dialogue: Support for deep multi-round conversation interactions
Natural Language Processing: Advanced language understanding capabilities based on DeepSeek API

System Requirements

Go 1.24.1 or higher
DeepSeek API key

Installation

git clone https://github.com/HundunOnline/mcp-agent-forge.git
cd agent-forge && make build

MCP Configuration

{
  "mcpServers": {
    "mcp-agent-forge": {
      "command": "/path/to/mcp-agent-forge",
      "env": {
        "DEEPSEEK_API_KEY": "xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxx",
      }
    }
  }
}

Configuration Guide

Configuration

Environment Variables

Variable Name	Description	Default Value	Required
`DEEPSEEK_API_KEY`	DeepSeek API Key	-	Yes
`LOG_LEVEL`	Logging level (debug, info, warn, error)	info	No
`LOG_PATH`	Log file path	./logs	No
`CONFIG_PATH`	Configuration file path	./config/config.yaml	No
`PORT`	Service port	8080	No
`DEBUG`	Debug mode switch	false	No

Usage

expert_personality_generation: Create a new agent
agent_answer: Simulate agent responses
get_agent: Get agent information
list_agents: List all agents
delete_agent: Delete an agent

Examples

Basic Usage

// Create an agent
{
    "name": "expert_personality_generation",
    "arguments": {
        "agent_name": "Elon Musk Thinking Model",
        "core_traits": "Systems Thinking,First Principles,Engineering Mindset,Risk Management,Innovation"
    }
}

// Agent response
{
    "name": "agent_answer",
    "arguments": {
        "agent_id": "your_agent_id",
        "context": "What's your view on Tesla's development strategy?",
        "planned_rounds": 3,
        "current_round": 1,
        "need_more_rounds": false
    }
}

Real Application Case

We created a sample application in Claude AI that demonstrates how to use Agent Forge to create and manage expert agents:

Claude AI Demo

This example shows:

How to create agents with specific professional backgrounds
How to conduct multi-round dialogue interactions
How to utilize agents' expertise to solve problems
How to manage and adjust agent behavior

Contributing

We welcome Pull Requests or Issues to help improve this project. We especially welcome contributions in the following areas:

New agent models and traits
Performance optimizations
Documentation improvements
Bug fixes
New feature suggestions

License

This project is licensed under the MIT License. See the LICENSE file for details.

This server cannot be installed

security - not tested

license - permissive license

quality - not tested

How are these scores calculated?

remote-capable server

The server can be hosted and run remotely because it primarily relies on remote services or has no dependency on the local environment.

A powerful MCP tool for forging and managing AI agent personalities with customizable expert traits and behaviors. 一个可以由大模型控制并定制LLM Agent的MCP工具，可以低成本快速复刻类似于CO-STORM的multi-agent research。

Related MCP Servers

Postman Tool Generation MCP Server
giovannicocco
A
security
A
license
A
quality
An MCP server that generates AI agent tools from Postman collections and requests. This server integrates with the Postman API to convert API endpoints into type-safe code that can be used with various AI frameworks.
Last updated -
1
7
JavaScript
MIT License
Tavily MCP Serverofficial
tavily-ai
A
security
A
license
A
quality
This server enables AI systems to integrate with Tavily's search and data extraction tools, providing real-time web information access and domain-specific searches.
Last updated -
2
8,610
494
JavaScript
MIT License
mcp-dingdingbot-server
HundunOnline
-
security
A
license
-
quality
dingding webhook mcp server
Last updated -
18
Go
GPL 3.0
MCP Prompt Tester
rt96-hub
-
security
A
license
-
quality
An MCP server that allows agents to test and compare LLM prompts across OpenAI and Anthropic models, supporting single tests, side-by-side comparisons, and multi-turn conversations.
Last updated -
Python
MIT License

View all related MCP servers

mcp-agent-forge

Agent Forge - 智能体锻造工具 (AI Agent Forge Tool)

中文版

功能特点

系统要求

安装

MCP 配置

配置说明

Configuration

Environment Variables

使用方法

示例

基本用法

实际应用案例

贡献指南

许可证

English Version

Features

System Requirements

Installation

MCP Configuration

Configuration Guide

Configuration

Environment Variables

Usage

Examples

Basic Usage

Real Application Case

Contributing

License

Related MCP Servers

Postman Tool Generation MCP Server

Tavily MCP Serverofficial

mcp-dingdingbot-server

MCP Prompt Tester

New MCP Servers

MCP directory API