ragflow-workbench

## 🇨🇳 中文介绍 **RAGFlow Workbench** 是一个面向 Windows 平台的 RAGFlow 全流程自动化技能包，覆盖从零搭建到日常运维的完整生命周期。 > 无需手动操作 Web 界面，不必记住复杂的 API 调用。一行命令，开箱即用。 ### 亮点 - **🚀 零到一全流程** — 从 Docker 安装检查、管理员注册、API Key 生成、模型配置到知识库管理的端到端自动化 - **🪟 Windows 原生优化** — 专为 Windows 环境设计，自动检测 Docker Desktop 和 RAGFlow 容器状态 - **📦 命令行驱动** — 全部操作通过 `uv run python scripts/*.py --json` 一键执行，结构化输出方便集成 - **🔐 安全优先** — 删除操作强制二次确认；.env 自动写入凭据，后续会话自动跳过安装引导 - **📊 完整可观测** — 解析进度实时反馈，FAIL 状态自动引导排查，不编造错误信息 ### 覆盖能力 | 模块 | 能力 | |------|------| | 环境检查 | Docker / 容器 / API / Web UI 四维健康报告 | | 管理员引导 | 自动注册、登录、JWT 获取、API Key 创建 | | 模型配置 | Embedding / Chat / Rerank 一键设置 | | 知识库 CRUD | 创建、查询、更新、删除（安全确认） | | 文档管理 | 上传、列表、更新、删除、批量操作 | | 解析生命周期 | 发起 / 停止 / 状态追踪 / 异常排查 | | 知识检索 | 语义检索、多库混合、KG 增强、rerank | | 对话创建 | 绑定知识库 + 对话模型，自定义提示词 | --- ## 🇬🇧 English Introduction **RAGFlow Workbench** is an end-to-end automation skill for the RAGFlow platform on Windows, covering the full lifecycle from first-time setup to daily knowledge base operations. > No manual Web UI navigation. No memorizing API payloads. One command from your terminal — ready to go. ### Highlights - **🚀 Full Lifecycle Automation** — From Docker installation check, admin registration, API key generation, model configuration to knowledge base management — all automated - **🪟 Windows-Native** — Designed specifically for Windows; automatically detects Docker Desktop, RAGFlow containers, and port availability - **📦 CLI-Driven** — Every operation runs via `uv run python scripts/*.py --json` for structured, pipable output - **🔐 Safety First** — Destructive operations require explicit confirmation; `.env` persists credentials so subsequent sessions skip setup - **📊 Full Observability** — Parse progress reported in real-time; FAIL states include actual error messages with guided troubleshooting ### Capabilities | Module | What It Does | |--------|-------------| | Environment Check | 4-dimension health report: Docker, container, API, Web UI | | Admin Bootstrap | Automatic registration, login, JWT & API key generation | | Model Setup | One-click Embedding / Chat / Rerank model configuration | | Knowledge Base CRUD | Create, inspect, update, delete (with safety confirmation) | | Document Management | Upload, list, update, delete, batch operations | | Parse Lifecycle | Start / stop / track / troubleshoot document parsing | | Knowledge Retrieval | Semantic search, cross-dataset, hybrid, KG-enhanced, rerank | | Chat Creation | Bind KB + LLM with customizable system prompts |