Dify Enterprise

https://enterprise-docs.dify.ai
Admin
Dify Enterprise is a private deployment AI middleware solution for large organizations and teams,...
Tokens:301,258
Snippets:2,007
Trust Score:9.9
Update:10 months ago
Show doc for...
Tokens:
Raw
### Install Python Dependencies for Dify Plugin

Source: https://enterprise-docs.dify.ai/plugin-dev-ja/9231-extension-plugin

This command installs the necessary Python packages required for the Dify AI plugin development, including Werkzeug for request handling, Flask for the web framework, and the dify-plugin library for plugin integration.

```bash
pip install werkzeug
pip install flask
pip install dify-plugin
```

--------------------------------

### 复制 `.env.example` 为 `.env`

Source: https://enterprise-docs.dify.ai/plugin_dev_zh/0211-getting-started-dify-tool

在插件项目根目录下，复制 `.env.example` 文件为 `.env`。这个 `.env` 文件将用于存储本地运行插件所需的 Dify 环境配置信息，如 API 主机和密钥。

```bash
cp .env.example .env
```

--------------------------------

### Get AKS Cluster Credentials (Bash)

Source: https://enterprise-docs.dify.ai/en-us/deployment/cloud-infrastructure/azure-poc-infra-setup

Retrieves the credentials for accessing an Azure Kubernetes Service (AKS) cluster using Azure CLI. This command configures kubectl to connect to the specified cluster. It requires the resource group and cluster name to be defined.

```bash
az aks get-credentials --resource-group $DIFY_RESOURCE_GROUP_NAME --name $DIFY_AKS_CLUSTER_NAME
```

--------------------------------

### 移动到插件项目上一级目录

Source: https://enterprise-docs.dify.ai/plugin_dev_zh/0211-getting-started-dify-tool

在执行打包命令前，需要将当前工作目录切换到插件项目文件夹的上一级目录。

```bash
cd ..

```

--------------------------------

### Agent Execution with Logging Example (Python)

Source: https://enterprise-docs.dify.ai/plugin_dev_zh/9232-agent

Demonstrates a two-step agent execution process using logging. It first yields a 'Thinking' log message with 'START' status, then performs an LLM invocation, and finally yields the completed log message with 'SUCCESS' status before yielding the actual text response.

```python
class FunctionCallingAgentStrategy(AgentStrategy):
    def _invoke(self, parameters: dict[str, Any]) -> Generator[AgentInvokeMessage]:
        thinking_log = self.create_log_message(
            data={
                "Query": parameters.get("query"),
            },
            label="Thinking",
            status=AgentInvokeMessage.LogMessage.LogStatus.START,
        )

        yield thinking_log

        llm_response = self.session.model.llm.invoke(
            model_config=LLMModelConfig(
                provider="openai",
                model="gpt-4o-mini",
                mode="chat",
                completion_params={},
            ),
            prompt_messages=[
                SystemPromptMessage(content="you are a helpful assistant"),
                UserPromptMessage(content=parameters.get("query")),
            ],
            stream=False,
            tools=[],
        )

        thinking_log = self.finish_log_message(
            log=thinking_log,
        )

        yield thinking_log

        yield self.create_text_message(text=llm_response.message.content)
```

--------------------------------

### Globally Install Dify Plugin CLI

Source: https://enterprise-docs.dify.ai/plugin_dev_zh/0131-cheatsheet

Installs the Dify Plugin CLI globally by moving the executable to a system path. This method is recommended for convenient access from any terminal. Example shown for macOS ARM.

```bash
# Rename and move to system path
# Example (macOS ARM)
mv dify-plugin-darwin-arm64 dify
sudo mv dify /usr/local/bin/
dify version
```

--------------------------------

### Python: 安装项目依赖

Source: https://enterprise-docs.dify.ai/plugin_dev_zh/0211-getting-started-dify-tool

该命令用于安装 `requirements.txt` 文件中列出的 Python 依赖项。这通常在激活虚拟环境后执行，以确保插件所需库被正确安装。

```bash
pip install -r requirements.txt
```

--------------------------------

### Python: 配置虚拟环境 (Windows)

Source: https://enterprise-docs.dify.ai/plugin_dev_zh/0211-getting-started-dify-tool

本代码块展示了在 Windows 系统上创建和激活 Python 虚拟环境的命令（适用于 cmd.exe 和 PowerShell）。这有助于隔离项目依赖，推荐在插件开发前执行。

```bash
venv\Scripts\activate.bat
# 或者
venv\Scripts\Activate.ps1
```

--------------------------------

### Initialize Dify Plugin Project

Source: https://enterprise-docs.dify.ai/plugin-dev-ja/9231-extension-plugin

Command-line instructions to create a new Dify plugin project using the Dify plugin scaffold tool. Ensure the tool is downloaded and executable in your path.

```bash
./dify-plugin-darwin-arm64 plugin init
```

```bash
dify plugin init
```

--------------------------------

### Execute IRSA One-Click Script for AWS Setup

Source: https://enterprise-docs.dify.ai/versions/3-0-x/zh-cn/deployment/cloud-infrastructure/aws-setup

This command executes a bash script named 'irsa_one_click.sh'. This script is designed for demonstration and testing purposes to automate the setup of IAM Roles for Service Accounts (IRSA) on AWS for Dify Enterprise Edition.

```bash
./irsa_one_click.sh
```

--------------------------------

### 打包 Dify 插件

Source: https://enterprise-docs.dify.ai/plugin_dev_zh/0211-getting-started-dify-tool

执行此命令来打包您的 Dify 插件项目。它会将插件的所有必需文件和配置编译成一个 `.difypkg` 文件，用于分发或安装。

```bash
./dify plugin package ./telegraph
```

--------------------------------

### Complete Agent Log Message (Python)

Source: https://enterprise-docs.dify.ai/plugin_dev_zh/9232-agent

Completes a previously started log message by updating its status. This function is used to change the status from 'START' to 'SUCCESS' or indicate an 'ERROR' if the task failed. It requires the log message object and optionally accepts an error message.

```python
def finish_log_message(
    self,
    log: AgentInvokeMessage,
    status: AgentInvokeMessage.LogMessage.LogStatus = AgentInvokeMessage.LogMessage.LogStatus.SUCCESS,
    error: Optional[str] = None,
) -> AgentInvokeMessage:
    # Implementation details omitted for brevity
    pass
```

--------------------------------

### Extract and Initialize Dify Enterprise Deployment

Source: https://enterprise-docs.dify.ai/versions/3-0-x/zh-cn/deployment/test-production/docker-compose

Extract the downloaded Dify Enterprise Docker Compose archive into a dedicated directory and then navigate into that directory to start the Docker Compose services.

```bash
mkdir dify-enterprise
tar -xzf dify-docker-compose-<version>.tgz -C ./dify-enterprise/
cd dify-enterprise
docker compose up -d
```

--------------------------------

### 配置 Dify 插件远程调试 (.env 文件)

Source: https://enterprise-docs.dify.ai/plugin-dev-ja/9231-extension-plugin

通过修改 `.env` 文件来配置 Dify 插件的远程调试。需要填写从“插件管理”页面获取的调试 Key 和远程服务器地址。该配置允许远程服务器连接并调试本地运行的插件。

```bash
INSTALL_METHOD=remote
REMOTE_INSTALL_URL=debug.dify.ai:5003
REMOTE_INSTALL_KEY=********-****-****-****-************
```

--------------------------------

### Start Qdrant Service with Docker Compose

Source: https://enterprise-docs.dify.ai/versions/3-0-x/zh-cn/deployment/advanced-configuration/qdrant/single-node

Launches the Qdrant service in detached mode using the previously configured Docker Compose file. This command assumes the docker-compose.yml file is present in the current directory.

```bash
docker-compose up -d
```

--------------------------------

### Python: 配置虚拟环境 (macOS/Linux)

Source: https://enterprise-docs.dify.ai/plugin_dev_zh/0211-getting-started-dify-tool

本代码块展示了在 macOS 或 Linux 系统上创建和激活 Python 虚拟环境的命令。这有助于隔离项目依赖，推荐在插件开发前执行。

```bash
python -m venv venv
source venv/bin/activate
```

--------------------------------

### 配置 `.env` 文件

Source: https://enterprise-docs.dify.ai/plugin_dev_zh/0211-getting-started-dify-tool

编辑 `.env` 文件，填入您的 Dify 实例的 API 主机地址和 API 密钥。这是本地运行插件服务时，插件与 Dify 环境通信所必需的。

```dotenv
DIFY_API_HOST=https://your-dify-host.com
DIFY_API_KEY=your-api-key
```

--------------------------------

### Get Dify Enterprise Web App Callback URL for External Users (Bash)

Source: https://enterprise-docs.dify.ai/versions/2-8-x/zh-cn/administrator-guide/sso/okta/oauth2

This snippet shows the Callback URL for Dify Enterprise Web App authentication for external users. This URL is required for the Okta application setup.

```bash
https://[your-app-url]/api/enterprise/sso/oauth2/callback
```

--------------------------------

### Python: Example of Invoking a Tool in an Agent Strategy

Source: https://enterprise-docs.dify.ai/plugin_dev_zh/9232-agent

Demonstrates how to use `self.session.tool.invoke()` within a `FunctionCallingAgentStrategy`. It retrieves tool details, constructs parameters, and calls the invoke method. The tool call name and arguments are typically generated by an LLM.

```python
from dify_plugin.entities.tool import ToolProviderType

class FunctionCallingAgentStrategy(AgentStrategy):
    def _invoke(self, parameters: dict[str, Any]) -> Generator[AgentInvokeMessage]:
        """
        Run FunctionCall agent application
        """
        fc_params = FunctionCallingParams(**parameters)
        
        # tool_call_name and tool_call_args parameter is obtained from the output of LLM
        tool_instances = {tool.identity.name: tool for tool in fc_params.tools} if fc_params.tools else {}
        tool_instance = tool_instances[tool_call_name]
        tool_invoke_responses = self.session.tool.invoke(
            provider_type=ToolProviderType.BUILT_IN,
            provider=tool_instance.identity.provider,
            tool_name=tool_instance.identity.name,
            # add the default value
            parameters={**tool_instance.runtime_parameters, **tool_call_args},
        )
```

--------------------------------

### 运行 Dify 插件进行调试 (Python)

Source: https://enterprise-docs.dify.ai/plugin-dev-ja/9231-extension-plugin

使用 Python 命令 `python -m main` 来启动已配置好远程调试的 Dify 插件。启动后，插件将显示在 Dify 的插件管理页面，并且其他团队成员也可以访问。此步骤是验证插件功能的基础。

```python
python -m main
```

--------------------------------

### 打包 Dify 插件 (命令行)

Source: https://enterprise-docs.dify.ai/plugin_dev_zh/0222-tool-plugin

使用 Dify CLI 工具打包插件。将 `./google` 替换为插件项目的实际路径，执行此命令后，会在当前文件夹生成 `google.difypkg` 文件，即最终的插件包。

```bash
# 将 ./google 替换为插件项目的实际路径

dify plugin package ./google
```

--------------------------------

### Backup Dify Community Edition Data and Configuration

Source: https://enterprise-docs.dify.ai/versions/3-0-x/zh-cn/deployment/test-production/docker-compose

Create a compressed tar archive containing the volumes and .env file of your Dify community edition installation. Include any other custom directories to ensure a complete backup before upgrading to the enterprise version.

```bash
tar -czf backup.tgz volumes .env  # 如有其他自定义文件夹请一并备份
```

--------------------------------

### 打包 Dify 插件 (CLI)

Source: https://enterprise-docs.dify.ai/plugin-dev-ja/9231-extension-plugin

使用 Dify CLI 工具 `dify plugin package` 命令来打包插件项目。该命令会生成一个 `.difypkg` 文件，即最终的插件包。请将 `./neko` 替换为插件项目的实际路径。

```bash
# 将 ./neko 替换为插件项目的实际路径

dify plugin package ./neko
```

--------------------------------

### Configure Third-Party Service Credentials (YAML)

Source: https://enterprise-docs.dify.ai/plugin_dev_zh/0222-tool-plugin

Integrate with third-party services by defining their required credentials within the provider's YAML configuration. This example shows how to add a `serpapi_api_key` for the SerpApi service, specifying its type, requirements, and display information for the user interface.

```yaml
identity:
    author: Dify
    name: google
    label:
        en_US: Google
        zh_Hans: Google
        pt_BR: Google
    description:
        en_US: Google
        zh_Hans: GoogleSearch
        pt_BR: Google
    icon: icon.svg
    tags:
        - search
credentials_for_provider:
    serpapi_api_key:
        type: secret-input
        required: true
        label:
            en_US: SerpApi API key
            zh_Hans: SerpApi API key
        placeholder:
            en_US: Please input your SerpApi API key
            zh_Hans: 请输入你的 SerpApi API key
        help:
            en_US: Get your SerpApi API key from SerpApi
            zh_Hans: 从 SerpApi 获取您的 SerpApi API key
        url: https://serpapi.com/manage-api-key
tools:
    - tools/google_search.yaml
extra:
    python:
        source: google.py

```

--------------------------------

### 实现 Dify Provider 凭证验证 (Python)

Source: https://enterprise-docs.dify.ai/plugin_dev_zh/0211-getting-started-dify-tool

此 Python 代码实现了 `ToolProvider` 类的 `_validate_credentials` 方法，用于验证 Telegraph 访问令牌。它首先检查令牌是否存在，然后尝试使用该令牌创建一个测试页面来验证其有效性。如果令牌缺失或 API 调用失败，将抛出 `ToolProviderCredentialValidationError` 异常，并附带详细的错误信息。

```python
from typing import Any
from dify_plugin import ToolProvider
from dify_plugin.errors.tool import ToolProviderCredentialValidationError

class TelegraphProvider(ToolProvider):
    def _validate_credentials(self, credentials: dict[str, Any]) -> None:
        """
        验证提供的 Telegraph Access Token 是否有效。
        尝试使用该 token 创建一个测试页面。
        如果验证失败，应抛出 ToolProviderCredentialValidationError 异常。
        """
        access_token = credentials.get("telegraph_access_token")
        if not access_token:
            raise ToolProviderCredentialValidationError("Telegraph Access Token 不能为空。")

        try:
            # 尝试执行一个需要凭证的简单操作来验证
            from ytelegraph import TelegraphAPI
            ph = TelegraphAPI(access_token=access_token)
            # 尝试创建一个临时的、无害的页面作为验证手段
            # 注意：更好的验证方式可能是调用 API 的 'getAccountInfo' 等只读方法（如果存在）
            test_page = ph.create_page_md("Dify Validation Test", "This is a test page created by Dify plugin validation.")
            # 如果需要，可以考虑立即编辑或删除这个测试页面，但这会增加复杂性
            # print(f"Validation successful. Test page created: {test_page}")
        except Exception as e:
            # 如果 API 调用失败，说明凭证很可能无效
            raise ToolProviderCredentialValidationError(f"Telegraph 凭证验证失败: {e}")

```
Tokens:
Raw