autogpt-agents

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

AutoGPT - Autonomous AI Agent Platform

AutoGPT - 自主AI Agent平台

Comprehensive platform for building, deploying, and managing continuous AI agents through a visual interface or development toolkit.

通过可视化界面或开发工具包构建、部署和管理持续运行AI Agent的综合平台。

When to use AutoGPT

何时使用AutoGPT

Use AutoGPT when:

Building autonomous agents that run continuously
Creating visual workflow-based AI agents
Deploying agents with external triggers (webhooks, schedules)
Building complex multi-step automation pipelines
Need a no-code/low-code agent builder

Key features:

Visual Agent Builder: Drag-and-drop node-based workflow editor
Continuous Execution: Agents run persistently with triggers
Marketplace: Pre-built agents and blocks to share/reuse
Block System: Modular components for LLM, tools, integrations
Forge Toolkit: Developer tools for custom agent creation
Benchmark System: Standardized agent performance testing

Use alternatives instead:

LangChain/LlamaIndex: If you need more control over agent logic
CrewAI: For role-based multi-agent collaboration
OpenAI Assistants: For simple hosted agent deployments
Semantic Kernel: For Microsoft ecosystem integration

在以下场景使用AutoGPT：

构建持续运行的自主Agent
创建基于可视化工作流的AI Agent
部署带有外部触发器（Webhook、定时任务）的Agent
构建复杂的多步骤自动化流水线
需要无代码/低代码的Agent构建器

核心功能：

可视化Agent构建器：基于节点的拖拽式工作流编辑器
持续执行：Agent可通过触发器持久运行
市场：可共享/复用的预构建Agent和组件块
组件块系统：用于LLM、工具、集成的模块化组件
Forge工具包：用于自定义Agent创建的开发者工具
基准测试系统：标准化的Agent性能测试

可选择替代方案的场景：

LangChain/LlamaIndex：如果需要对Agent逻辑有更多控制
CrewAI：用于基于角色的多Agent协作
OpenAI Assistants：用于简单的托管式Agent部署
Semantic Kernel：用于微软生态系统集成

Quick start

快速开始

Installation (Docker)

安装（Docker方式）

bash

undefined

bash

undefined

Clone repository

git clone https://github.com/Significant-Gravitas/AutoGPT.git cd AutoGPT/autogpt_platform

Copy environment file

cp .env.example .env

Start backend services

docker compose up -d --build

Start frontend (in separate terminal)

cd frontend cp .env.example .env npm install npm run dev

undefined

cd frontend cp .env.example .env npm install npm run dev

undefined

Access the platform

访问平台

Frontend UI: http://localhost:3000
Backend API: http://localhost:8006/api
WebSocket: ws://localhost:8001/ws

前端UI：http://localhost:3000
后端API：http://localhost:8006/api
WebSocket：ws://localhost:8001/ws

Architecture overview

架构概述

AutoGPT has two main systems:

AutoGPT包含两个主要系统：

AutoGPT Platform (Production)

AutoGPT平台（生产环境）

Visual agent builder with React frontend
FastAPI backend with execution engine
PostgreSQL + Redis + RabbitMQ infrastructure

基于React的可视化Agent构建器前端
带有执行引擎的FastAPI后端
PostgreSQL + Redis + RabbitMQ基础设施

AutoGPT Classic (Development)

AutoGPT经典版（开发环境）

Forge: Agent development toolkit
Benchmark: Performance testing framework
CLI: Command-line interface for development

Forge：Agent开发工具包
Benchmark：性能测试框架
CLI：用于开发的命令行界面

Core concepts

核心概念

Graphs and nodes

图与节点

Agents are represented as graphs containing nodes connected by links:

Graph (Agent)
  ├── Node (Input)
  │   └── Block (AgentInputBlock)
  ├── Node (Process)
  │   └── Block (LLMBlock)
  ├── Node (Decision)
  │   └── Block (SmartDecisionMaker)
  └── Node (Output)
      └── Block (AgentOutputBlock)

Agent以图的形式呈现，包含由链接连接的节点：

Graph (Agent)
  ├── Node (Input)
  │   └── Block (AgentInputBlock)
  ├── Node (Process)
  │   └── Block (LLMBlock)
  ├── Node (Decision)
  │   └── Block (SmartDecisionMaker)
  └── Node (Output)
      └── Block (AgentOutputBlock)

Blocks

组件块

Blocks are reusable functional components:

Block Type	Purpose
`INPUT`	Agent entry points
`OUTPUT`	Agent outputs
`AI`	LLM calls, text generation
`WEBHOOK`	External triggers
`STANDARD`	General operations
`AGENT`	Nested agent execution

组件块是可复用的功能组件：

组件块类型	用途
`INPUT`	Agent入口点
`OUTPUT`	Agent输出端
`AI`	LLM调用、文本生成
`WEBHOOK`	外部触发器
`STANDARD`	通用操作
`AGENT`	嵌套Agent执行

Execution flow

执行流程

User/Trigger → Graph Execution → Node Execution → Block.execute()
     ↓              ↓                 ↓
  Inputs      Queue System      Output Yields

用户/触发器 → 图执行 → 节点执行 → Block.execute()
     ↓              ↓                 ↓
  输入值      队列系统      输出结果

Building agents

构建Agent

Using the visual builder

使用可视化构建器

Open Agent Builder at http://localhost:3000
Add blocks from the BlocksControl panel
Connect nodes by dragging between handles
Configure inputs in each node
Run agent using PrimaryActionBar

打开位于http://localhost:3000的**Agent构建器**
从BlocksControl面板添加组件块
通过拖拽手柄连接节点
在每个节点中配置输入
使用PrimaryActionBar运行Agent

Available blocks

可用组件块

AI Blocks:

```
AITextGeneratorBlock
```
- Generate text with LLMs
```
AIConversationBlock
```
- Multi-turn conversations
```
SmartDecisionMakerBlock
```
- Conditional logic

Integration Blocks:

GitHub, Google, Discord, Notion connectors
Webhook triggers and handlers
HTTP request blocks

Control Blocks:

Input/Output blocks
Branching and decision nodes
Loop and iteration blocks

AI组件块：

```
AITextGeneratorBlock
```
- 利用LLM生成文本
```
AIConversationBlock
```
- 多轮对话
```
SmartDecisionMakerBlock
```
- 条件逻辑

集成组件块：

GitHub、Google、Discord、Notion连接器
Webhook触发器和处理器
HTTP请求组件块

控制组件块：

输入/输出组件块
分支和决策节点
循环和迭代组件块

Agent execution

Agent执行

Trigger types

触发器类型

Manual execution:

http

POST /api/v1/graphs/{graph_id}/execute
Content-Type: application/json

{
  "inputs": {
    "input_name": "value"
  }
}

Webhook trigger:

http

POST /api/v1/webhooks/{webhook_id}
Content-Type: application/json

{
  "data": "webhook payload"
}

Scheduled execution:

json

{
  "schedule": "0 */2 * * *",
  "graph_id": "graph-uuid",
  "inputs": {}
}

手动执行：

http

POST /api/v1/graphs/{graph_id}/execute
Content-Type: application/json

{
  "inputs": {
    "input_name": "value"
  }
}

Webhook触发：

http

POST /api/v1/webhooks/{webhook_id}
Content-Type: application/json

{
  "data": "webhook payload"
}

定时执行：

json

{
  "schedule": "0 */2 * * *",
  "graph_id": "graph-uuid",
  "inputs": {}
}

Monitoring execution

监控执行

WebSocket updates:

javascript

const ws = new WebSocket('ws://localhost:8001/ws');

ws.onmessage = (event) => {
  const update = JSON.parse(event.data);
  console.log(`Node ${update.node_id}: ${update.status}`);
};

REST API polling:

http

GET /api/v1/executions/{execution_id}

WebSocket更新：

javascript

const ws = new WebSocket('ws://localhost:8001/ws');

ws.onmessage = (event) => {
  const update = JSON.parse(event.data);
  console.log(`Node ${update.node_id}: ${update.status}`);
};

REST API轮询：

http

GET /api/v1/executions/{execution_id}

Using Forge (Development)

使用Forge（开发环境）

Create custom agent

创建自定义Agent

bash

undefined

bash

undefined

Setup forge environment

cd classic ./run setup

Create new agent from template

./run forge create my-agent

Start agent server

./run forge start my-agent

undefined

./run forge start my-agent

undefined

Agent structure

Agent结构

my-agent/
├── agent.py          # Main agent logic
├── abilities/        # Custom abilities
│   ├── __init__.py
│   └── custom.py
├── prompts/          # Prompt templates
└── config.yaml       # Agent configuration

my-agent/
├── agent.py          # 主Agent逻辑
├── abilities/        # 自定义能力
│   ├── __init__.py
│   └── custom.py
├── prompts/          # 提示词模板
└── config.yaml       # Agent配置

Implement custom ability

实现自定义能力

python

from forge import Ability, ability

@ability(
    name="custom_search",
    description="Search for information",
    parameters={
        "query": {"type": "string", "description": "Search query"}
    }
)
def custom_search(query: str) -> str:
    """Custom search ability."""
    # Implement search logic
    result = perform_search(query)
    return result

python

from forge import Ability, ability

@ability(
    name="custom_search",
    description="Search for information",
    parameters={
        "query": {"type": "string", "description": "Search query"}
    }
)
def custom_search(query: str) -> str:
    """Custom search ability."""
    # 实现搜索逻辑
    result = perform_search(query)
    return result

Benchmarking agents

Agent基准测试

Run benchmarks

运行基准测试

bash

undefined

bash

undefined

Run all benchmarks

运行所有基准测试

./run benchmark

Run specific category

运行特定分类

./run benchmark --category coding

Run with specific agent

使用特定Agent运行

./run benchmark --agent my-agent

undefined

./run benchmark --agent my-agent

undefined

Benchmark categories

基准测试分类

Coding: Code generation and debugging
Retrieval: Information finding
Web: Web browsing and interaction
Writing: Text generation tasks

Coding：代码生成与调试
Retrieval：信息查找
Web：网页浏览与交互
Writing：文本生成任务

VCR cassettes

VCR录像带

Benchmarks use recorded HTTP responses for reproducibility:

bash

undefined

基准测试使用录制的HTTP响应以确保可复现性：

bash

undefined

Record new cassettes

录制新的录像带

./run benchmark --record

Run with existing cassettes

使用现有录像带运行

./run benchmark --playback

undefined

./run benchmark --playback

undefined

Integrations

集成

Adding credentials

添加凭证

Navigate to Profile > Integrations
Select provider (OpenAI, GitHub, Google, etc.)
Enter API keys or authorize OAuth
Credentials are encrypted and stored securely

导航至个人资料 > 集成
选择提供商（OpenAI、GitHub、Google等）
输入API密钥或授权OAuth
凭证将被加密并安全存储

Using credentials in blocks

在组件块中使用凭证

Blocks automatically access user credentials:

python

class MyLLMBlock(Block):
    def execute(self, inputs):
        # Credentials are injected by the system
        credentials = self.get_credentials("openai")
        client = OpenAI(api_key=credentials.api_key)
        # ...

组件块会自动访问用户凭证：

python

class MyLLMBlock(Block):
    def execute(self, inputs):
        # 系统会注入凭证
        credentials = self.get_credentials("openai")
        client = OpenAI(api_key=credentials.api_key)
        # ...

Supported providers

支持的提供商

Provider	Auth Type	Use Cases
OpenAI	API Key	LLM, embeddings
Anthropic	API Key	Claude models
GitHub	OAuth	Code, repos
Google	OAuth	Drive, Gmail, Calendar
Discord	Bot Token	Messaging
Notion	OAuth	Documents

提供商	认证类型	使用场景
OpenAI	API密钥	LLM、嵌入向量
Anthropic	API密钥	Claude模型
GitHub	OAuth	代码、仓库
Google	OAuth	云端硬盘、Gmail、日历
Discord	机器人令牌	消息通讯
Notion	OAuth	文档

Deployment

部署

Docker production setup

Docker生产环境配置

yaml

undefined

yaml

undefined

docker-compose.prod.yml

services: rest_server: image: autogpt/platform-backend environment: - DATABASE_URL=postgresql://... - REDIS_URL=redis://redis:6379 ports: - "8006:8006"

executor: image: autogpt/platform-backend command: poetry run executor

frontend: image: autogpt/platform-frontend ports: - "3000:3000"

undefined

services: rest_server: image: autogpt/platform-backend environment: - DATABASE_URL=postgresql://... - REDIS_URL=redis://redis:6379 ports: - "8006:8006"

executor: image: autogpt/platform-backend command: poetry run executor

frontend: image: autogpt/platform-frontend ports: - "3000:3000"

undefined

Environment variables

环境变量

Variable	Purpose
`DATABASE_URL`	PostgreSQL connection
`REDIS_URL`	Redis connection
`RABBITMQ_URL`	RabbitMQ connection
`ENCRYPTION_KEY`	Credential encryption
`SUPABASE_URL`	Authentication

变量	用途
`DATABASE_URL`	PostgreSQL连接地址
`REDIS_URL`	Redis连接地址
`RABBITMQ_URL`	RabbitMQ连接地址
`ENCRYPTION_KEY`	凭证加密密钥
`SUPABASE_URL`	认证服务地址

Generate encryption key

生成加密密钥

bash

cd autogpt_platform/backend
poetry run cli gen-encrypt-key

bash

cd autogpt_platform/backend
poetry run cli gen-encrypt-key

Best practices

最佳实践

Start simple: Begin with 3-5 node agents
Test incrementally: Run and test after each change
Use webhooks: External triggers for event-driven agents
Monitor costs: Track LLM API usage via credits system
Version agents: Save working versions before changes
Benchmark: Use agbenchmark to validate agent quality

从简开始：先构建包含3-5个节点的Agent
增量测试：每次修改后运行并测试
使用Webhook：通过外部触发器实现事件驱动型Agent
监控成本：通过额度系统跟踪LLM API使用情况
版本化Agent：修改前保存可用版本
基准测试：使用agbenchmark验证Agent质量

Common issues

常见问题

Services not starting:

bash

undefined

服务无法启动：

bash

undefined

Check container status

检查容器状态

docker compose ps

View logs

查看日志

docker compose logs rest_server

Restart services

重启服务

docker compose restart


**Database connection issues:**
```bash

docker compose restart


**数据库连接问题：**
```bash

Run migrations

运行迁移

cd backend poetry run prisma migrate deploy


**Agent execution stuck:**
```bash

cd backend poetry run prisma migrate deploy


**Agent执行停滞：**
```bash

Check RabbitMQ queue

检查RabbitMQ队列

Visit http://localhost:15672 (guest/guest)

访问http://localhost:15672（账号/密码：guest/guest）

Clear stuck executions

清除停滞的执行任务

docker compose restart executor

undefined

docker compose restart executor

undefined

References

参考资料

Advanced Usage - Custom blocks, deployment, scaling
Troubleshooting - Common issues, debugging

高级用法 - 自定义组件块、部署、扩容
故障排除 - 常见问题、调试方法

Resources

资源

Documentation: https://docs.agpt.co
Repository: https://github.com/Significant-Gravitas/AutoGPT
Discord: https://discord.gg/autogpt
License: MIT (Classic) / Polyform Shield (Platform)

文档：https://docs.agpt.co
代码仓库：https://github.com/Significant-Gravitas/AutoGPT
Discord社区：https://discord.gg/autogpt
许可证：MIT（经典版）/ Polyform Shield（平台版）