经 AI Skill Hub 精选评估,AI工作流平台 获评「强烈推荐」。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 8.0 分,适合有一定技术背景的用户使用。
自托管AI平台,支持推理、工具使用、浏览器自动化、图像生成等
AI工作流平台 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。
自托管AI平台,支持推理、工具使用、浏览器自动化、图像生成等
AI工作流平台 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。
# 克隆仓库 git clone https://github.com/psyb0t/aigate cd aigate # 查看安装说明 cat README.md # 按 README 完成环境依赖安装后即可使用
# 查看帮助 aigate --help # 基本运行 aigate [options] <input> # 详细使用说明请查阅文档 # https://github.com/psyb0t/aigate
# aigate 配置说明 # 查看配置选项 aigate --config-example > config.yml # 常见配置项 # output_dir: ./output # log_level: info # workers: 4 # 环境变量(覆盖配置文件) export AIGATE_CONFIG="/path/to/config.yml"
A self-hosted AI platform. One docker-compose up.
Everything an AI-powered workflow needs — inference, tool use, browser automation, image generation, speech synthesis, transcription, object storage, agentic code execution, web search, an email gateway, a Telegram client, time-series forecasting, an async job queue, and a web UI — behind a single OpenAI-compatible endpoint at http://localhost:4000. Point any existing client at it and it works.
MCP tools across multiple servers. Any model with function calling can invoke them autonomously — the model orchestrates, you just prompt.
voice field), Qwen3-TTS with voice cloning (CUDA), OpenAI TTS.search_web tool./telethon/. Send/read/edit/delete messages, list dialogs, forward, send files from URL, manage group membership. REST API + MCP tools./mailbox/. Stateless IMAP+SMTP across N accounts from one YAML config — unified inbox, per-account list/search/CRUD, SMTP send. REST API + flat MCP tool set (mailbox parameter selects account)./predictalot/. Five foundation forecasters (chronos-2, timesfm-2.5, moirai-2, toto-1, sundial-base-128m) across six forecast types (univariate, multivariate, past/future covariates, samples) with per-type weighted ensembles. REST API + 26 MCP tools. CPU or CUDA.Ask a Groq model to research something and it opens a browser, reads pages, screenshots them, uploads to storage, and comes back with a summary and links. The model decides what tools to use and in what order.
```bash
curl http://localhost:4000/predictalot/v1/univariate/forecast \ -H "Authorization: Bearer $PREDICTALOT_AUTH_TOKEN" \ -H "Content-Type: application/json" \ -d '{ "model": "chronos-2", "context": [[10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20]], "config": {"horizon": 5} }'
cp .env.example .env
Fill in the values — every variable is documented with comments in .env.example.
Everything is opt-in via flags in .env. API keys are stored separately and never activate anything on their own — set the flag to 1 to enable:
| Flag | What it enables |
|---|---|
OPENAI=1 | OpenAI models (gpt-4o, o3, DALL-E, Whisper, TTS) |
ANTHROPIC=1 | Direct Anthropic API models |
CLAUDEBOX=1 | claudebox service + models + MCP server (Claude Code via OAuth or API key) |
PIBOX_ZAI=1 | pibox-zai service + GLM models + MCP server (pi-coding-agent via z.ai) |
CEREBRAS=1 | Cerebras models (free: 5 RPM / 30K TPM / 1M TPD, 4 models only — see [limits](docs/providers.md#free-tier-reality-check)) |
OPENROUTER=1 | OpenRouter models (free: 50 RPD at $0, 1K RPD at $10+ credits — see [limits](docs/providers.md#free-tier-reality-check)) |
HUGGINGFACE=1 | HuggingFace models (free: **$0.10/mo credits only**, eval tier — see [limits](docs/providers.md#free-tier-reality-check)) |
MISTRAL=1 | Mistral AI models (free "Experiment" tier — exact limits not published, see your Admin dashboard) |
COHERE=1 | Cohere models (trial: **1K calls/MONTH total cap** — see [limits](docs/providers.md#free-tier-reality-check)) |
GROQ=1 | Groq models (free: 30 RPM, 1K-14.4K RPD per model — see [limits](docs/providers.md#free-tier-reality-check)) |
OLLAMA=1 | Local Ollama CPU inference (~6GB+ RAM) |
OLLAMA_CUDA=1 | Local Ollama NVIDIA GPU inference (requires nvidia-container-toolkit) |
SPEACHES=1 | Local Speaches CPU transcription/TTS (~4GB RAM) |
SPEACHES_CUDA=1 | Local CUDA-accelerated STT (requires nvidia-container-toolkit) |
QWEN_TTS_CUDA=1 | Local CUDA-accelerated TTS via Qwen3 (voice cloning, requires nvidia-container-toolkit) |
SDCPP=1 | Local stable-diffusion.cpp CPU image generation |
SDCPP_CUDA=1 | Local stable-diffusion.cpp CUDA image generation (requires nvidia-container-toolkit) |
HYBRIDS3=1 | Object storage service + MCP server (S3-compatible, plain HTTP, auto-expiry) |
BROWSER=1 | Stealth browser cluster + MCP server (~1.3GB RAM) |
LIBRECHAT=1 | LibreChat web UI at /librechat/ with all models and MCP tools |
SEARXNG=1 | SearXNG meta-search at /searxng/ + MCP search_web tool |
TELETHON=1 | Telegram client at /telethon/ + MCP Telegram tools |
CLOUDFLARED=1 | Cloudflare Tunnel |
TAILSCALE=1 | Tailscale node — tailnet-only HTTP proxy to nginx (no public exposure) |
PREDICTALOT=1 | predictalot at /predictalot/ — time-series forecasting + MCP (CPU image) |
PREDICTALOT_CUDA=1 | predictalot CUDA variant (requires nvidia-container-toolkit) |
MAILBOX=1 | mailbox IMAP+SMTP gateway at /mailbox/ + MCP (needs MAILBOX_CONFIG + MAILBOX_AUTH_TOKEN) |
make run regenerates litellm/config.yaml before starting — only enabled providers are included, fallback chains are filtered to match.
If CLOUDFLARED_CONFIG or CLOUDFLARED_CREDS are set, make run/make run-bg will verify those files exist before starting. A missing file causes an immediate error — better than Docker silently mounting a directory instead.
curl "http://localhost:4000/mailbox/inbox?limit=5" \ -H "Authorizati
高质量的AI工作流平台,支持多种AI任务
该工具未明确声明开源协议,商业使用前请联系原作者确认授权范围,避免侵权风险。
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。
AI Skill Hub 点评:AI工作流平台 的核心功能完整,质量优秀。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 | aigate |
| 原始描述 | 开源AI工作流:A self-hosted AI platform — inference, tool use, browser automation, image gener。⭐8 · Shell |
| Topics | aianthropiccerebrasclaudeshell |
| GitHub | https://github.com/psyb0t/aigate |
| 语言 | Shell |
收录时间:2026-05-27 · 更新时间:2026-05-27 · License:未公布 · 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