AI工作流自动化 是 AI Skill Hub 本期精选Agent工作流之一。综合评分 8.0 分,整体质量较高。我们强烈推荐将其纳入你的 AI 工具库,帮助提升工作效率。
AI工作流自动化 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。
AI工作流自动化 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 dqiii8
# 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 dqiii8
# 方式三:从源码安装(获取最新功能)
git clone https://github.com/senda-labs/DQIII8
cd DQIII8
pip install -e .
# 验证安装
python -c "import dqiii8; print('安装成功')"
# 命令行使用
dqiii8 --help
# 基本用法
dqiii8 input_file -o output_file
# Python 代码中调用
import dqiii8
# 示例
result = dqiii8.process("input")
print(result)
# dqiii8 配置文件示例(config.yml) app: name: "dqiii8" debug: false log_level: "INFO" # 运行时指定配置文件 dqiii8 --config config.yml # 或通过环境变量配置 export DQIII8_API_KEY="your-key" export DQIII8_OUTPUT_DIR="./output"
<p align="center"> <h1 align="center">DQIII8</h1> <p align="center">Autonomous, Cost-First Multi-Agent Orchestration Engine</p> <p align="center"> <img alt="Tests" src="https://img.shields.io/badge/tests-1004%20passing-brightgreen"> <img alt="License: MIT" src="https://img.shields.io/badge/license-MIT-blue.svg"> <img alt="Python 3.10+" src="https://img.shields.io/badge/python-3.10%2B-blue"> <img alt="Platform" src="https://img.shields.io/badge/platform-Ubuntu%2022.04%2F24.04-lightgrey"> <img alt="Claude Code" src="https://img.shields.io/badge/Claude%20Code-integrated-blueviolet"> </p> </p>
DQIII8 is an autonomous AI orchestration engine built for SSH-only VPS deployment. Every request flows through a cost-first routing pipeline that always tries the cheapest capable model first — local → free cloud → paid frontier — escalating only when the task demands it. It is deeply integrated with Claude Code through 15 lifecycle hooks, 22 skills, and 17 specialist agents. (Counts are validator-enforced against the live tree — check_readme_counts() in bin/tools/validate_rules_registry.py; CLAUDE.md:16 is the canonical restatement.)
This repository is a reference implementation. It shows the architecture, routing logic, hook system, and agent patterns so you can build a similar system with your own models and providers. The knowledge base, databases, and credentials are populated locally — see Installation.
---
Requirements: Ubuntu 22.04/24.04 (or WSL2), Python 3.10+. Ollama is optional (enables Tier C local models). The knowledge/RAG layer is also optional.
git clone https://github.com/senda-labs/DQIII8
cd DQIII8
bash install.sh
The installer: 1. Installs Python dependencies (hash-verified from requirements.lock if present, via pip install --require-hashes; falls back to unpinned requirements.txt otherwise). Regenerate the lock after editing requirements.txt with: pip install pip-tools && pip-compile --generate-hashes --output-file=requirements.lock requirements.txt 2. Prompts to install Ollama (optional — skip to start with Tier B only) 3. Pulls qwen2.5-coder:7b if Ollama is installed 4. Copies config/.env.example → .env if not present 5. Applies database/schema_v2.sql (creates all 60 tables) 6. Copies Claude Code settings template 7. Runs smoke tests
Enable the optional knowledge base (requires Ollama + bge-m3):
bash install.sh --with-knowledge This additionally pulls bge-m3, indexes the 5 knowledge domains, seeds domain classifier centroids, and migrates embeddings to sqlite-vec.
python3 -m pytest tests/test_smoke.py -q
```bash
Copy config/.env.example to .env:
| Variable | Tier | Required | Source |
|---|---|---|---|
GROQ_API_KEY | B | **Yes (free)** | [console.groq.com](https://console.groq.com) |
NVIDIA_API_KEY | B+ | Recommended (free) | [integrate.api.nvidia.com](https://integrate.api.nvidia.com) |
GITHUB_TOKEN | B++ | Optional (free) | GitHub → Settings → Developer settings |
ANTHROPIC_API_KEY | A/S | Optional (paid) | [console.anthropic.com](https://console.anthropic.com) |
TELEGRAM_BOT_TOKEN | UI | Optional | [@BotFather](https://t.me/BotFather) |
OLLAMA_BASE_URL | C | Optional | default: http://localhost:11434 |
Claude Code OAuth is also supported for Anthropic calls — set ANTHROPIC_API_KEY="" in subprocess env to force OAuth instead of the direct API key.
---
nano .env
高效的AI工作流自动化工具
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。
✅ MIT 协议 — 最宽松的开源协议之一,可自由商用、修改、分发,仅需保留版权声明。
经综合评估,AI工作流自动化 在Agent工作流赛道中表现稳健,质量优秀。如果你已有明确的使用需求,可以直接上手体验;如果还在评估阶段,建议对比同类工具后再做决策。
| 原始名称 | DQIII8 |
| 原始描述 | 开源AI工作流:Works for you. Go outside and live. — AI orchestrator that auto-routes tasks to 。⭐10 · Python |
| Topics | ai-agentai-automationai-orchestration |
| GitHub | https://github.com/senda-labs/DQIII8 |
| License | MIT |
| 语言 | Python |
收录时间:2026-05-31 · 更新时间:2026-05-31 · License:MIT · 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