# anp-examples

**Repository Path**: gjunming/anp-examples

## Basic Information

- **Project Name**: anp-examples
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: MIT
- **Default Branch**: deploy_demo
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2025-07-21
- **Last Updated**: 2025-07-30

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# ANP网络探索工具 / ANP Network Explorer

[English](#english) | [中文](#chinese)

<a name="chinese"></a>
## 中文文档

### 环境设置

在运行项目之前，你需要设置必要的环境变量。项目中提供了一个 `.env.example` 文件作为模板：

1.  复制 `.env.example` 文件并重命名为 `.env`：
    ```bash
    cp .env.example .env
    ```
2.  编辑 `.env` 文件，填入你的 Azure OpenAI API 密钥和 Endpoint 等实际配置信息。

### 项目介绍

ANP网络探索工具是一个基于Agent Network Protocol (ANP)的应用程序，允许用户使用自然语言与智能体网络进行交互。用户可以提供智能体描述URL，通过简单的问题与智能体进行对话，并实时查看网络爬取的过程。

### 项目结构

本项目包含以下主要组件：

- **web_app/**：Web应用程序，包含前端和后端实现
  - **frontend/**：基于HTML/JavaScript的用户界面
  - **backend/**：基于FastAPI的后端服务器
  - **static/**：静态资源文件

- **anp_examples/**：ANP核心功能实现
  - **simple_example.py**：简化的ANP爬取逻辑实现
  - **anp_tool.py**：ANP工具类，用于与智能体网络交互
  - **utils/**：工具类和辅助函数

- **use_did_test_public/**：DID认证相关文件
  - **did.json**：DID文档
  - **key-1_private.pem**：私钥文件
  - **private_keys.json**：密钥配置

- **examples_code/**：示例代码
  - **client.py**：客户端示例
  - **server.py**：服务器示例
  - **did_auth_middleware.py**：DID认证中间件
  - **jwt_config.py**：JWT配置

### 如何使用

#### 使用Web应用程序

1. 安装依赖：
   ```bash
   # 使用Poetry
   poetry install
   
   # 或使用pip
   pip install -r web_app/backend/requirements.txt
   ```

2. 启动Web应用程序：
   ```bash
   # 使用Poetry
   ./web_app/run_with_poetry.sh
   
   # 或使用脚本
   ./web_app/run.sh
   ```

3. 打开浏览器访问：`http://localhost:8000`

4. 在输入框中输入您的问题，并提供智能体URL（可选，默认为`https://agent-search.ai/ad.json`）

5. 点击"提交问题"按钮，查看结果和网络爬取过程

#### 使用ANP示例

您也可以直接使用`anp_examples`中的代码：

```python
from anp_examples.simple_example import simple_crawl

# 使用simple_crawl函数
result = await simple_crawl(
    user_input="您的问题",
    task_type="general",
    initial_url="https://agent-search.ai/ad.json"  # 智能体URL
)

# 查看结果
print(result["content"])  # 回答内容
print(result["visited_urls"])  # 访问过的URL
```

### 开发

1. 克隆仓库：
   ```bash
   git clone https://github.com/yourusername/anp-examples.git
   cd anp-examples
   ```

2. 安装开发依赖：
   ```bash
   poetry install
   ```

3. 运行测试：
   ```bash
   pytest
   ```

---

<a name="english"></a>
## English Documentation

### Environment Setup

Before running the project, you need to set up the necessary environment variables. An `.env.example` file is provided as a template:

1.  Copy the `.env.example` file and rename it to `.env`:
    ```bash
    cp .env.example .env
    ```
2.  Edit the `.env` file and fill in your actual Azure OpenAI API Key, Endpoint, and other configuration details.

### Project Introduction

ANP Network Explorer is an application based on the Agent Network Protocol (ANP) that allows users to interact with agent networks using natural language. Users can provide agent description URLs, engage in conversations with agents through simple questions, and view the network crawling process in real-time.

### Project Structure

This project contains the following main components:

- **web_app/**: Web application, including frontend and backend implementations
  - **frontend/**: HTML/JavaScript-based user interface
  - **backend/**: FastAPI-based backend server
  - **static/**: Static resource files

- **anp_examples/**: ANP core functionality implementation
  - **simple_example.py**: Simplified ANP crawling logic
  - **anp_tool.py**: ANP tool class for interacting with agent networks
  - **utils/**: Utility classes and helper functions

- **use_did_test_public/**: DID authentication related files
  - **did.json**: DID document
  - **key-1_private.pem**: Private key file
  - **private_keys.json**: Key configuration

- **examples_code/**: Example code
  - **client.py**: Client example
  - **server.py**: Server example
  - **did_auth_middleware.py**: DID authentication middleware
  - **jwt_config.py**: JWT configuration

### How to Use

#### Using the Web Application

1. Install dependencies:
   ```bash
   # Using Poetry
   poetry install
   
   # Or using pip
   pip install -r web_app/backend/requirements.txt
   ```

2. Start the web application:
   ```bash
   # Using Poetry
   ./web_app/run_with_poetry.sh
   
   # Or using script
   ./web_app/run.sh
   ```

3. Open browser and visit: `http://localhost:8000`

4. Enter your question in the input box and provide an agent URL (optional, default is `https://agent-search.ai/ad.json`)

5. Click the "Submit" button to view results and the network crawling process

#### Using ANP Examples

You can also directly use the code in `anp_examples`:

```python
from anp_examples.simple_example import simple_crawl

# Use the simple_crawl function
result = await simple_crawl(
    user_input="Your question",
    task_type="general",
    initial_url="https://agent-search.ai/ad.json"  # Agent URL
)

# View results
print(result["content"])  # Answer content
print(result["visited_urls"])  # Visited URLs
```

### Development

1. Clone the repository:
   ```bash
   git clone https://github.com/yourusername/anp-examples.git
   cd anp-examples
   ```

2. Install development dependencies:
   ```bash
   poetry install
   ```

3. Run tests:
   ```bash
   pytest
   ```

# anp-examples
anp-examples

Develop an ANP example application

It consists of two parts:

1. **ANP Agent**
The entry point of the agent is an agent description document. Through this document, connections to internal agent data can be established. The agent description document, combined with internal data such as additional JSON files, images, and interface files, constitutes the public information of the agent. It is recommended to use a hotel agent as an example. Construct the agent's data, including a hotel description, services provided by the hotel, customer service details, and booking interfaces. Use FastAPI to return the relevant documents based on requests. I can provide sample documents for the hotel agent.

2. **ANP Client**
Develop a client that accesses the ANP agent. The client will feature a page that accepts a URL pointing to an agent description document. With this document URL, the client can access all information from the agent, including services, products, and API endpoints like hotel booking interfaces. The page should clearly display which URLs the client accessed and the content retrieved, allowing users to visually follow the interaction process.