经 AI Skill Hub 精选评估,Vantyx AI 获评「推荐使用」。这款Agent工作流在功能完整性、社区活跃度和易用性方面表现出色,AI 评分 7.5 分,适合有一定技术背景的用户使用。
Vantyx AI 是一套完整的 AI Agent 自动化工作流方案。通过可视化的节点编排,将复杂的多步骤任务拆解为清晰的自动化流程,实现全程无人值守的智能处理。支持与数百种外部服务和 API 无缝集成,适合构建数据处理管线、业务自动化和 AI 辅助决策系统。
Vantyx AI 是一套完整的 AI Agent 自动化工作流方案。通过可视化的节点编排,将复杂的多步骤任务拆解为清晰的自动化流程,实现全程无人值守的智能处理。支持与数百种外部服务和 API 无缝集成,适合构建数据处理管线、业务自动化和 AI 辅助决策系统。
# 方式一:pip 安装(推荐)
pip install vantyx-ai
# 方式二:虚拟环境安装(推荐生产环境)
python -m venv .venv
source .venv/bin/activate # Windows: .venv\Scripts\activate
pip install vantyx-ai
# 方式三:从源码安装(获取最新功能)
git clone https://github.com/felipebridge/vantyx-ai
cd vantyx-ai
pip install -e .
# 验证安装
python -c "import vantyx_ai; print('安装成功')"
# 命令行使用
vantyx-ai --help
# 基本用法
vantyx-ai input_file -o output_file
# Python 代码中调用
import vantyx_ai
# 示例
result = vantyx_ai.process("input")
print(result)
# vantyx-ai 配置文件示例(config.yml) app: name: "vantyx-ai" debug: false log_level: "INFO" # 运行时指定配置文件 vantyx-ai --config config.yml # 或通过环境变量配置 export VANTYX_AI_API_KEY="your-key" export VANTYX_AI_OUTPUT_DIR="./output"
<p><strong>Agent infrastructure for teams that ship to production.</strong></p> <p> Build, deploy, and operate autonomous AI agents and multi-agent workflows at scale.<br/> Full lifecycle management — from creation and memory to orchestration and observability. </p>
<br/>
<p> <img src="https://img.shields.io/badge/agent%20runtime-ReAct%20loop-1AB5C8?style=flat-square&logoColor=white" /> <img src="https://img.shields.io/badge/providers-OpenAI%20%C2%B7%20Anthropic%20%C2%B7%20Local-3EC97A?style=flat-square&logoColor=white" /> <img src="https://img.shields.io/badge/storage-PostgreSQL%20%2B%20Redis-7DC843?style=flat-square&logoColor=white" /> </p>
<p> <a href="https://github.com/vantyx-ai/vantyx/actions/workflows/ci.yml"> <img src="https://img.shields.io/github/actions/workflow/status/vantyx-ai/vantyx/ci.yml?label=CI&style=flat-square&color=1AB5C8" /> </a> <a href="https://github.com/vantyx-ai/vantyx/actions/workflows/security.yml"> <img src="https://img.shields.io/github/actions/workflow/status/vantyx-ai/vantyx/security.yml?label=security&style=flat-square&color=3EC97A" /> </a> <a href="https://codecov.io/gh/vantyx-ai/vantyx"> <img src="https://img.shields.io/codecov/c/github/vantyx-ai/vantyx?style=flat-square&color=7DC843" /> </a> <a href="https://pypi.org/project/vantyx/"> <img src="https://img.shields.io/pypi/v/vantyx?style=flat-square&color=1AB5C8" /> </a> <a href="https://pypi.org/project/vantyx/"> <img src="https://img.shields.io/pypi/pyversions/vantyx?style=flat-square&color=3EC97A" /> </a> <a href="LICENSE"> <img src="https://img.shields.io/badge/license-Apache%202.0-7DC843?style=flat-square" /> </a> </p>
<br/> </div>
<br/>
---
Most agent frameworks are prototypes dressed as infrastructure. Vantyx is built the other way around: PostgreSQL-backed persistence, async-first I/O, provider-agnostic completions, full execution tracing, and YAML-defined workflows that compose into production pipelines.
It handles the operational complexity so your team can focus on agent behavior.
---
```bash git clone https://github.com/vantyx-ai/vantyx cd vantyx cp .env.example .env
docker compose up -d
docker compose exec app alembic upgrade head
Multi-platform images published to GHCR on each release:
docker pull ghcr.io/vantyx-ai/vantyx:0.1.0
Kubernetes manifests: infra/k8s/
---
| Example | Description |
|---|---|
[examples/basic_agent.py](examples/basic_agent.py) | Create and run an agent, retrieve the execution trace |
[examples/multi_agent_workflow.py](examples/multi_agent_workflow.py) | Parallel agents + synthesis in a single workflow |
[examples/memory_continuity.py](examples/memory_continuity.py) | Multi-turn conversation with resumable sessions |
[examples/custom_tool.py](examples/custom_tool.py) | Register tools and use them from an agent |
[examples/workflows/research_pipeline.yaml](examples/workflows/research_pipeline.yaml) | Full YAML workflow definition |
---
make up make migrate ```
API live at http://localhost:8000. Interactive docs at /docs (when DEBUG=true).
All settings via environment variables (.env):
| Variable | Description | Default |
|---|---|---|
SECRET_KEY | JWT signing key (≥ 32 chars) | required |
DATABASE_URL | PostgreSQL async DSN | postgresql+asyncpg://... |
REDIS_URL | Redis DSN | redis://localhost:6379/0 |
ENV | development / staging / production | development |
OPENAI_API_KEY | OpenAI API key | optional |
ANTHROPIC_API_KEY | Anthropic API key | optional |
LOCAL_MODEL_ENDPOINT | OpenAI-compatible local model endpoint | optional |
MAX_AGENT_STEPS | ReAct loop step limit | 50 |
LOG_FORMAT | json or text | json |
DEBUG | Enable /docs and verbose errors | false |
---
Workflows compose agents, tools, conditions, and transforms into DAGs:
version: "1.0"
steps:
- type: parallel
name: research
steps:
- type: agent
name: technical
agent_id: "uuid"
message: "Technical analysis of {{topic}}"
output_key: tech
- type: agent
name: market
agent_id: "uuid"
message: "Market analysis of {{topic}}"
output_key: market
- type: condition
name: depth_check
condition: detail_level == "deep"
branches:
"true":
- type: agent
name: deep_dive
agent_id: "uuid"
message: "Deep dive:\n\nTech: {{tech}}\n\nMarket: {{market}}"
output_key: report
"false":
- type: transform
name: combine
mapping:
report: "{{tech}}"
```bash
高质量的开源AI工作流框架,值得关注
该工具使用 NOASSERTION 协议,商用场景请仔细阅读协议条款,必要时咨询法律意见。
AI Skill Hub 为第三方内容聚合平台,本页面信息基于公开数据整理,不对工具功能和质量作任何法律背书。
建议在沙箱或测试环境中充分验证后,再部署至生产环境,并做好必要的安全评估。
📄 NOASSERTION — 请查阅原始协议条款了解具体使用限制。
AI Skill Hub 点评:Vantyx AI 的核心功能完整,质量良好。对于自动化工程师和运维人员来说,这是一个值得纳入个人工具库的选择。建议先在非生产环境试用,再逐步推广。
| 原始名称 | vantyx-ai |
| 原始描述 | 开源AI工作流:AI agent runtime and orchestration framework for autonomous workflows, persisten。⭐5 · Python |
| Topics | AI工作流自动化 |
| GitHub | https://github.com/felipebridge/vantyx-ai |
| License | NOASSERTION |
| 语言 | Python |
收录时间:2026-05-28 · 更新时间:2026-05-30 · License:NOASSERTION · AI Skill Hub 不对第三方内容的准确性作法律背书。
选择 Agent 类型,复制安装指令后粘贴到对应客户端