Documentation
Complete guide to using Codemus for visual Agile project orchestration with AI agents.
🚀 Getting Started
Learn the basics and get up and running quickly.
📚 Core Concepts
Understand the fundamental concepts behind Codemus.
⚙️ Features
Explore all the features and capabilities of Codemus.
🔧 Advanced Topics
Deep dive into advanced features and configurations.
What is Agile Vibecoding?
Agile Vibecoding is a flow state where you orchestrate AI agents through visual, real-time interactions. Instead of breaking your focus to manage tasks, switch contexts, or track progress, Codemus keeps you in the zone with:
- Visual Flow: See your entire project as a living, interactive graph that updates in real-time
- Voice-First Interaction: Speak commands naturally without leaving your coding flow
- Contextual Conversations: Per-task conversations that maintain context without switching screens
- Autonomous Orchestration: PM Agent handles coordination so you can focus on what matters
- Zero Context Switching: Everything you need is visible in one unified graph view
What is Codemus?
Codemus is an Agent Operating System — a macOS desktop application that enables Agile Vibecoding through AI-powered agent orchestration. Unlike traditional project management tools, Codemus provides a runtime environment where a Project Manager (PM) Agent coordinates Specialist Agents (Developers, Designers, QA Engineers, etc.) to execute tasks autonomously, all visualized in real-time.
Key Features
- Agile Vibecoding: Maintain flow state with visual orchestration and voice-first interaction
- Visual Orchestration: Real-time interactive graph showing all entities and their relationships
- Agile Enforcement: Hard-gate state machine ensuring proper project lifecycle
- Event Sourcing: Every action creates an immutable event for complete auditability
- Time Travel: Replay to any point in history, create alternate timelines
- Tool Governance: 4-tier ring system controlling what tools agents can access
- PM Autonomy: Configurable autonomy levels from manual to fully automatic
- Document RAG: Import documents and make them searchable for context
- Voice Input: On-device speech recognition for hands-free operation