Merge branch 'main' into wegent/team-icon-favorites-showcase

Author: qdaxb

.github/workflows/test.yml

@@ -64,6 +64,7 @@ jobs:
   test-executor:
     name: Test Executor
     runs-on: ubuntu-latest
+    timeout-minutes: 10
 
     steps:
       - name: Checkout code

AGENTS.md

@@ -26,7 +26,7 @@ Wegent is an open-source AI-native operating system for defining, organizing, an
 **Multi-module architecture:**
 - **Backend** (FastAPI + SQLAlchemy + MySQL): RESTful API and business logic
 - **Frontend** (Next.js 15 + TypeScript + React 19): Web UI with shadcn/ui components
-- **Executor**: Task execution engine (Claude Code, Agno)
+- **Executor**: Task execution engine (Claude Code, Agno, Dify)
 - **Executor Manager**: Task orchestration via Docker
 - **Shared**: Common utilities and models
 
@@ -465,13 +465,21 @@ alembic upgrade head --sql
 
 ### Executor
 
-**Tech:** Python, Claude Code SDK, Agno, Docker
+**Tech:** Python, Claude Code SDK, Agno, Dify API, Docker
+
+**Supported Agent Types:**
+- **ClaudeCode**: For code development tasks with Claude Code SDK
+- **Agno**: For dialogue and chat tasks
+- **Dify**: For external API integration with Dify platform (chat, workflow, chatflow, agent-chat modes)
 
 **Common tasks:**
 - Add agent type: Implement in `agents/`
 - Modify execution: Update `tasks/`
 
-**Environment:** `ANTHROPIC_AUTH_TOKEN` (Claude Code) or `ANTHROPIC_API_KEY` (Agno)
+**Environment Variables:**
+- Claude Code: `ANTHROPIC_AUTH_TOKEN`
+- Agno: `ANTHROPIC_API_KEY`
+- Dify: `DIFY_API_KEY`, `DIFY_BASE_URL`
 
 ### Executor Manager
 

README.md

@@ -160,6 +160,7 @@ graph LR
 
    - **Claude Code Shell**: Uses `ANTHROPIC_AUTH_TOKEN`
    - **Agno Shell**: Uses `ANTHROPIC_API_KEY`
+   - **Dify Shell**: Uses `DIFY_API_KEY` and `DIFY_BASE_URL`
 
    Please set the correct variable based on your Shell configuration. Check the Shell's documentation or the `executor/agents/` code for specific requirements.
 
@@ -190,11 +191,11 @@ graph TB
     
     subgraph "🤖 Agent Layer"
         Claude["🧠 Claude Code"]
-        Ango["💻 Agno"]
-        DifyPlanned["✨ Dify (Planned)"]
+        Agno["💻 Agno"]
+        Dify["✨ Dify"]
     end
-  
-    
+
+
     %% System Interactions
     Frontend --> API
     API --> Backend
@@ -203,11 +204,11 @@ graph TB
     ExecutorManager --> Executor1
     ExecutorManager --> Executor2
     ExecutorManager --> ExecutorN
-    
-    %% AI Program Integration (Currently only supports Claude Code)
+
+    %% AI Program Integration
     Executor1 --> Claude
-    Executor2 --> Claude
-    ExecutorN --> Ango
+    Executor2 --> Agno
+    ExecutorN --> Dify
 ```
 
 ## 🛠️ Development

README_zh.md

@@ -162,6 +162,7 @@ graph LR
 
    - **Claude Code Shell**: 使用 `ANTHROPIC_AUTH_TOKEN`
    - **Agno Shell**: 使用 `ANTHROPIC_API_KEY`
+   - **Dify Shell**: 使用 `DIFY_API_KEY` 和 `DIFY_BASE_URL`
 
    请根据您的 Shell 配置设置正确的环境变量。具体要求请查看 Shell 文档或 `executor/agents/` 代码。
 
@@ -192,8 +193,8 @@ graph TB
 
     subgraph "🤖 智能体层"
         Claude["🧠 Claude Code"]
-        Ango["💻 Agno"]
-        DifyPlanned["✨ Dify（计划中）"]
+        Agno["💻 Agno"]
+        Dify["✨ Dify"]
     end
 
 
@@ -206,10 +207,10 @@ graph TB
     ExecutorManager --> Executor2
     ExecutorManager --> ExecutorN
 
-    %% AI 程序集成（目前仅支持 Claude Code）
+    %% AI 程序集成
     Executor1 --> Claude
-    Executor2 --> Claude
-    ExecutorN --> Ango
+    Executor2 --> Agno
+    ExecutorN --> Dify
 ```
 
 ## 🛠️ 开发

docker-compose.yml

@@ -45,7 +45,7 @@ services:
   backend:
     # build:
       # dockerfile: docker/backend/Dockerfile
-    image: ghcr.io/wecode-ai/wegent-backend:1.0.11
+    image: ghcr.io/wecode-ai/wegent-backend:1.0.12
     container_name: wegent-backend
     restart: always
     ports:
@@ -79,7 +79,7 @@ services:
   frontend:
     # build:
     #   dockerfile: docker/frontend/Dockerfile
-    image: ghcr.io/wecode-ai/wegent-web:1.0.11
+    image: ghcr.io/wecode-ai/wegent-web:1.0.12
     container_name: wegent-frontend
     restart: always
     ports:
@@ -96,7 +96,7 @@ services:
   executor_manager:
     #build:
       #dockerfile: docker/frontend/Dockerfile
-    image: ghcr.io/wecode-ai/wegent-executor-manager:1.0.11
+    image: ghcr.io/wecode-ai/wegent-executor-manager:1.0.12
     extra_hosts:
       - "host.docker.internal:host-gateway"
     container_name: wegent-executor-manager
@@ -109,7 +109,7 @@ services:
       - PORT=8001
       - CALLBACK_HOST=http://executor_manager:8001
       - NETWORK=wegent-network
-      - EXECUTOR_IMAGE=ghcr.io/wecode-ai/wegent-executor:1.0.11
+      - EXECUTOR_IMAGE=ghcr.io/wecode-ai/wegent-executor:1.0.12
       - EXECUTOR_PORT_RANGE_MIN=10001
       - EXECUTOR_PORT_RANGE_MAX=10100
       - EXECUTOR_WORKSPCE=${HOME}/wecode-bot

docs/en/concepts/architecture.md

@@ -44,7 +44,7 @@ graph TB
     subgraph "🤖 Agent Layer"
         Claude["🧠 Claude Code<br/>Coding Agent"]
         Agno["💻 Agno<br/>Chat Agent"]
-        DifyPlanned["✨ Dify<br/>Planned"]
+        Dify["✨ Dify<br/>External API Agent"]
     end
 
 
@@ -60,8 +60,8 @@ graph TB
 
     %% AI Program Integration
     Executor1 --> Claude
-    Executor2 --> Claude
-    ExecutorN --> Agno
+    Executor2 --> Agno
+    ExecutorN --> Dify
 
     %% Styling
     classDef platform fill:#e3f2fd,stroke:#1976d2,stroke-width:2px
@@ -72,7 +72,7 @@ graph TB
     class Frontend,Backend,API platform
     class MySQL,Redis data
     class ExecutorManager,Executor1,Executor2,ExecutorN execution
-    class Claude,Agno,DifyPlanned agent
+    class Claude,Agno,Dify agent
 ```
 
 ### Architecture Layers
@@ -82,7 +82,7 @@ graph TB
 | **Management Platform Layer** | User interaction, resource management, API services | Next.js 15, FastAPI, React 19 |
 | **Data Layer** | Data persistence, cache management | MySQL 9.4, Redis 7 |
 | **Execution Layer** | Task scheduling, container orchestration, resource isolation | Docker, Python |
-| **Agent Layer** | AI capabilities, code execution, chat processing | Claude Code, Agno |
+| **Agent Layer** | AI capabilities, code execution, chat processing, external API integration | Claude Code, Agno, Dify |
 
 ---
 
@@ -206,7 +206,7 @@ EXECUTOR_IMAGE: wegent-executor:1.0.5 # Executor image
 
 **Technology Stack**:
 - **Container**: Docker
-- **Runtime**: Claude Code, Agno
+- **Runtime**: Claude Code, Agno, Dify
 - **Version Control**: Git
 
 **Core Features**:
@@ -389,8 +389,8 @@ container:
 
 executor_engines:
   - "Claude Code (Anthropic)"
-  - "Agno (experimental)"
-  - "Dify (planned)"
+  - "Agno"
+  - "Dify"
 ```
 
 ---

docs/en/guides/user/creating-bots.md

@@ -40,7 +40,7 @@ Bot = Ghost (soul) + Shell (container) + Model (AI model)
 | Component | Description | Example |
 |-----------|-------------|---------|
 | **Ghost** | Defines agent's personality and capabilities | "Frontend development expert" |
-| **Shell** | Runtime environment | ClaudeCode, Agno |
+| **Shell** | Runtime environment | ClaudeCode, Agno, Dify |
 | **Model** | AI model configuration | Claude Sonnet 4, GPT-4 |
 
 ### Bot vs Ghost
@@ -103,6 +103,7 @@ Clarify what this Bot will be used for:
 **Choosing Shell**:
 - ClaudeCode: Suitable for code development tasks
 - Agno: Suitable for dialogue and interaction tasks
+- Dify: Suitable for external API integration with Dify platform (supports chat, workflow, chatflow, agent-chat modes)
 
 **Choosing Model**:
 - Sonnet: Balance performance and cost
@@ -449,6 +450,9 @@ shellRef: ClaudeCode
 
 # Dialogue interaction tasks
 shellRef: Agno
+
+# External API integration with Dify
+shellRef: Dify
 ```
 
 ### 3. Reuse Strategy

docs/zh/concepts/architecture.md

@@ -44,7 +44,7 @@ graph TB
     subgraph "🤖 智能体层"
         Claude["🧠 Claude Code<br/>代码智能体"]
         Agno["💻 Agno<br/>对话智能体"]
-        DifyPlanned["✨ Dify<br/>计划中"]
+        Dify["✨ Dify<br/>外部 API 智能体"]
     end
 
 
@@ -60,8 +60,8 @@ graph TB
 
     %% AI 程序集成
     Executor1 --> Claude
-    Executor2 --> Claude
-    ExecutorN --> Agno
+    Executor2 --> Agno
+    ExecutorN --> Dify
 
     %% 样式
     classDef platform fill:#e3f2fd,stroke:#1976d2,stroke-width:2px
@@ -72,7 +72,7 @@ graph TB
     class Frontend,Backend,API platform
     class MySQL,Redis data
     class ExecutorManager,Executor1,Executor2,ExecutorN execution
-    class Claude,Agno,DifyPlanned agent
+    class Claude,Agno,Dify agent
 ```
 
 ### 架构层次说明
@@ -82,7 +82,7 @@ graph TB
 | **管理平台层** | 用户交互、资源管理、API 服务 | Next.js 15, FastAPI, React 19 |
 | **数据层** | 数据持久化、缓存管理 | MySQL 9.4, Redis 7 |
 | **执行层** | 任务调度、容器编排、资源隔离 | Docker, Python |
-| **智能体层** | AI 能力提供、代码执行、对话处理 | Claude Code, Agno |
+| **智能体层** | AI 能力提供、代码执行、对话处理、外部 API 集成 | Claude Code, Agno, Dify |
 
 ---
 
@@ -206,7 +206,7 @@ EXECUTOR_IMAGE: wegent-executor:1.0.5 # 执行器镜像
 
 **技术栈**：
 - **容器**: Docker
-- **运行时**: Claude Code, Agno
+- **运行时**: Claude Code, Agno, Dify
 - **版本控制**: Git
 
 **核心特性**：
@@ -389,8 +389,8 @@ container:
 
 executor_engines:
   - "Claude Code (Anthropic)"
-  - "Agno (experimental)"
-  - "Dify (planned)"
+  - "Agno"
+  - "Dify"
 ```
 
 ---

docs/zh/guides/user/creating-bots.md

@@ -40,7 +40,7 @@ Bot = Ghost (灵魂) + Shell (容器) + Model (AI 模型)
 | 组件 | 说明 | 示例 |
 |------|------|------|
 | **Ghost** | 定义智能体的个性和能力 | "前端开发专家" |
-| **Shell** | 运行时环境 | ClaudeCode, Agno |
+| **Shell** | 运行时环境 | ClaudeCode, Agno, Dify |
 | **Model** | AI 模型配置 | Claude Sonnet 4, GPT-4 |
 
 ### Bot vs Ghost
@@ -103,6 +103,7 @@ Bot 通过 `Ref` (引用) 来组合资源,而不是直接包含配置。这样
 **选择 Shell**:
 - ClaudeCode: 适合代码开发任务
 - Agno: 适合对话和交互任务
+- Dify: 适合与 Dify 平台的外部 API 集成（支持 chat、workflow、chatflow、agent-chat 模式）
 
 **选择 Model**:
 - Sonnet: 平衡性能和成本
@@ -449,6 +450,9 @@ shellRef: ClaudeCode
 
 # 对话交互任务
 shellRef: Agno
+
+# 与 Dify 外部 API 集成
+shellRef: Dify
 ```
 
 ### 3. 复用策略