# AGiXT — Extensible AI Agent Automation Framework

> Open-source AI agent automation platform with 50+ provider integrations, plugin system, chain-of-thought workflows, and persistent memory. Self-hostable via Docker.

## Install

Save as a script file and run:

## Quick Use

```bash
git clone https://github.com/Josh-XT/AGiXT
cd AGiXT
docker compose up -d
```

Open `http://localhost:8501` for the web UI.

## What is AGiXT?

AGiXT is an open-source AI agent automation platform that combines 50+ LLM providers, a plugin system for external tools, chain-of-thought workflows (called "chains"), and persistent memory. It provides a web UI, REST API, and Python SDK for building and managing autonomous AI agents.

**Answer-Ready**: AGiXT is an open-source AI agent automation platform with 50+ LLM providers, plugin-based tool system, chain-of-thought workflows, and persistent memory. Self-hostable via Docker with web UI and REST API. 3k+ GitHub stars.

**Best for**: Teams building customizable autonomous AI agents with tool integrations. **Works with**: Claude, GPT, Gemini, Mistral, Ollama, 45+ more providers. **Setup time**: Under 10 minutes with Docker.

## Core Features

### 1. Multi-Provider Support

| Category | Providers |
|----------|----------|
| Commercial | OpenAI, Anthropic, Google, Cohere, Mistral |
| Open Source | Ollama, vLLM, Hugging Face, llama.cpp |
| Cloud | Azure, AWS Bedrock, Google Vertex |
| Gateways | OpenRouter, LiteLLM, Portkey |

### 2. Plugin System
Extend agents with tools:

```
Built-in plugins:
- Web browsing and scraping
- Code execution (Python, Shell)
- File operations
- Email (SMTP, IMAP)
- GitHub integration
- Database queries (SQL)
- Image generation (DALL-E, SD)
- Voice (TTS, STT)
```

### 3. Chains (Multi-Step Workflows)

```yaml
Chain: "Research and Report"
Steps:
  1. Web Search: Search for {topic}
  2. Summarize: Condense search results
  3. Analyze: Extract key insights
  4. Generate: Write formatted report
  5. Email: Send report to {recipient}
```

### 4. Persistent Memory
Agents remember across sessions:
- Conversation history
- Learned facts and preferences
- Task outcomes
- Custom knowledge bases

### 5. REST API

```python
from agixtsdk import AGiXTSDK

agixt = AGiXTSDK(base_uri="http://localhost:7437")

# Create agent
agixt.add_agent("researcher", provider="anthropic", model="claude-sonnet-4-20250514")

# Chat
response = agixt.chat("researcher", "Research quantum computing trends")

# Run chain
result = agixt.run_chain("Research and Report", prompt="AI coding tools 2026")
```

### 6. Web UI
Streamlit-based interface for:
- Agent creation and configuration
- Chat interface
- Chain builder (visual)
- Memory management
- Plugin configuration

## Architecture

```
Web UI / REST API / Python SDK
            ↓
      AGiXT Core Engine
    ├── Agent Manager
    ├── Chain Executor
    ├── Memory (ChromaDB / PostgreSQL)
    ├── Plugin Loader
    └── Provider Router (50+ LLMs)
```

## FAQ

**Q: Is AGiXT free?**
A: Yes, fully open-source under MIT license. Self-host for free.

**Q: How does it compare to AutoGPT?**
A: AGiXT is more modular with its plugin/chain system and supports 50+ providers. AutoGPT is simpler but less extensible.

**Q: Can I use it in production?**
A: Yes, Docker deployment is production-ready with PostgreSQL backend.

## Source & Thanks

> Created by [Josh-XT](https://github.com/Josh-XT). Licensed under MIT.
>
> [Josh-XT/AGiXT](https://github.com/Josh-XT/AGiXT) — 3k+ stars

<!-- ZH -->

## 快速使用

```bash
git clone https://github.com/Josh-XT/AGiXT
cd AGiXT && docker compose up -d
```

Docker 一键启动，浏览器打开 `localhost:8501`。

## 什么是 AGiXT？

AGiXT 是开源 AI 代理自动化平台，50+ LLM 供应商、插件工具系统、链式工作流和持久记忆。

**一句话总结**：开源 AI 代理自动化平台，50+ LLM 供应商、插件系统、链式工作流、持久记忆，Docker 自托管。

**适合人群**：构建可定制自主 AI 代理的团队。

## 核心功能

### 1. 50+ 供应商
商业、开源、云端 LLM 全覆盖。

### 2. 插件系统
浏览器、代码执行、邮件、GitHub、数据库等。

### 3. 链式工作流
多步骤任务流，可视化构建。

### 4. 持久记忆
跨会话记忆对话、事实和任务结果。

## 常见问题

**Q: 免费吗？**
A: 完全开源免费，MIT 许可。

**Q: 和 AutoGPT 比较？**
A: AGiXT 更模块化，50+ 供应商，插件/链系统更可扩展。

## 来源与致谢

> [Josh-XT/AGiXT](https://github.com/Josh-XT/AGiXT) — 3k+ stars, MIT

---
Source: https://tokrepo.com/en/workflows/6528233f-b920-475b-8ecb-a15801303634
Author: Agent Toolkit