Agent Craft

v1.0.0Coding Agentsstable

AI Agent 教学仓库 | 系统化 LangChain、RAG、LangGraph、MCP 全栈实战代码 | 万字博客详解 | 开源可运行示例 | 从零构建智能体

ai-agentartificial-intelligencebeginner-guidechatbotdeepseek
Share:
314
Stars
0
Downloads
0
Weekly
0/5

What is Agent Craft?

Agent Craft is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to ai agent 教学仓库 | 系统化 langchain、rag、langgraph、mcp 全栈实战代码 | 万字博客详解 | 开源可运行示例 | 从零构建智能体

AI Agent 教学仓库 | 系统化 LangChain、RAG、LangGraph、MCP 全栈实战代码 | 万字博客详解 | 开源可运行示例 | 从零构建智能体

This server falls under the Coding Agents category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • AI Agent 教学仓库 | 系统化 LangChain、RAG、LangGraph、MCP 全栈实战代码 | 万字博

Use Cases

AI agent teaching repository
LangChain and LangGraph examples
Full-stack agent implementation
Annyfee

Maintainer

LicenseMIT
Languagepython
Versionv1.0.0
UpdatedMay 20, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

Manual Installation

npx agent-craft

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use Agent Craft

Agent Craft is a comprehensive open-source teaching repository that covers the full AI agent development stack through 13+ runnable modules — from basic LLM API calls and prompt engineering through LangChain, RAG with vector databases, LangGraph stateful workflows, multi-agent coordination, and MCP server implementation. It includes a hands-on Module 10 that teaches building private MCP servers using the FastMCP framework with both Stdio and Streamable HTTP protocols, plus Module 11 for client-side MCP integration. This is an ideal resource for developers who want structured, practical code examples rather than abstract documentation.

Prerequisites

  • Python 3.10 or higher installed
  • OpenAI or DeepSeek API key (set in .env file)
  • LangSmith account and API key for tracing (optional but recommended)
  • Git to clone the repository
  • An MCP-compatible client such as Claude Desktop for testing MCP modules
1

Clone the repository

Clone the Agent Craft repository from GitHub to your local machine and navigate into it.

git clone https://github.com/Annyfee/agent-craft.git
cd agent-craft
2

Install dependencies

Install all required Python packages. The project uses a pip-installable package structure so you can install it in editable mode along with its requirements.

pip install -r requirements.txt
pip install -e .
3

Configure environment variables

Copy the example environment file and fill in your API keys. At minimum you need an OpenAI or DeepSeek key; LangSmith is optional for tracing agent runs.

cp .env.example .env
# Edit .env and set:
# OPENAI_API_KEY=sk-...
# DEEPSEEK_API_KEY=...
# LANGSMITH_API_KEY=...
4

Run the introductory agent demo

Start with the first module to verify your setup is correct. This runs a basic agent demonstration showing LLM API calls and function calling.

python "m01_agent_introduction/Agent-demo.py"
5

Explore the MCP module

Navigate to Module 10 to learn how to build a private MCP server using the FastMCP framework. The module demonstrates both Stdio and Streamable HTTP communication protocols.

cd m10_mcp_basics
python server.py
6

Integrate MCP client with LangChain

Module 11 shows client-side MCP integration using the langchain-mcp-adapters framework, connecting your agent to any MCP server including the one built in Module 10.

cd m11_mcp_advanced
python client.py

Agent Craft Examples

Client configuration

Example MCP client configuration if connecting Claude Desktop to the FastMCP server built in Module 10.

{
  "mcpServers": {
    "agent-craft": {
      "command": "python",
      "args": ["m10_mcp_basics/server.py"],
      "env": {
        "OPENAI_API_KEY": "your_openai_api_key"
      }
    }
  }
}

Prompts to try

Example learning paths and prompts when using the Agent Craft modules as reference.

- "Show me how to implement a RAG pipeline with FAISS and LangChain"
- "Build a LangGraph workflow with human-in-the-loop approval"
- "Create a multi-agent system where agents hand off tasks to each other"
- "Demonstrate how to wrap an external API as an MCP tool using FastMCP"

Troubleshooting Agent Craft

ImportError or ModuleNotFoundError when running module scripts

Ensure you ran 'pip install -e .' from the repository root to install the package in editable mode. Also confirm you are using the correct Python environment with 'which python' or 'python --version'.

LLM API calls fail with authentication errors

Verify your .env file is in the repository root and contains the correct key names (OPENAI_API_KEY or DEEPSEEK_API_KEY). The project uses python-dotenv to load environment variables; confirm it is installed with 'pip show python-dotenv'.

Vector database or RAG modules fail during embedding

FAISS and Chroma require compatible versions of the underlying native libraries. Run 'pip install faiss-cpu' for CPU-only environments. If on Apple Silicon, use 'pip install faiss-cpu --no-binary faiss-cpu' or switch to the Chroma backend.

Frequently Asked Questions about Agent Craft

What is Agent Craft?

Agent Craft is a Model Context Protocol (MCP) server that ai agent 教学仓库 | 系统化 langchain、rag、langgraph、mcp 全栈实战代码 | 万字博客详解 | 开源可运行示例 | 从零构建智能体 It connects AI assistants to external tools and data sources through a standardized interface.

How do I install Agent Craft?

Follow the installation instructions on the Agent Craft GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.

Which AI clients work with Agent Craft?

Agent Craft works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is Agent Craft free to use?

Yes, Agent Craft is open source and available under the MIT license. You can use it freely in both personal and commercial projects.

Browse More Coding Agents MCP Servers

Explore all coding agents servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "agent-craft": { "command": "npx", "args": ["-y", "agent-craft"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use Agent Craft?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides