Libre Workspace AI Documentation

Welcome to the technical and conceptual documentation of Libre Workspace AI, an autonomous, modular, graph-based AI assistant system.

Concepts & Architecture

• General Architecture
  • Overview
  • The 4-Layer Conceptual Architecture
  • Centralized Context Builder (context_builder.py)
  • Shared State Model (storage)
  • Multi-User Isolation & OIDC Readiness
  • Loop Protection (Schleifenschutz)

Core Components

• Nodes
  • Overview
  • The Node Lifecycle & Interface
  • Built-in Node Categories
• Workflows
  • Overview
  • Workflow Structure & Graph Definition
  • Workflow Composition & Sub-Workflows
  • Workflow Types
  • Execution Engine & Control Flow
• Skills
  • Overview
  • 1. Python Domain Skills (Folder-Based Architecture)
  • 2. Prompt Skills (SKILL.md)
  • 3. MCP Skills (*.mcp)
  • 4. Configured Skill Instances
  • Dynamic Skill Evaluation Loop

Interfaces & Tooling

• CLI Reference
  • Overview
  • Starting the CLI
  • Interactive Slash Commands
  • Execution Lifecycle & Live Streaming
• WebGUI
  • Overview
  • Starting the WebGUI
  • Core Views & Features
• REST API Reference
  • Overview
  • Authentication & Multi-User Context
  • Core Endpoints

Development & Contributing

• Development Guidelines
  • Setup & Installation
  • Strict No-AI / No-LLM Testing Policy
  • Prompt Language & Anonymity Guidelines
  • Creating a Custom Node

Indices and tables

• Index

• Module Index

• Search Page