
What is klaw?
klaw is an open-source platform for deploying and managing AI agents at scale. Think of it as Kubernetes, but instead of containers, you’re orchestrating intelligent agents that can code, research, communicate, and automate tasks.Quick Start
Get up and running with klaw in under 5 minutes
Core Concepts
Understand the fundamental building blocks
Guides
Step-by-step tutorials for common use cases
CLI Reference
Complete command-line reference
Why klaw?
The Problem
AI agents are powerful, but deploying them is a mess:- Claude Code is great but single-user and closed-source
- LangChain/CrewAI require complex Python setups and don’t scale
- Custom solutions mean reinventing auth, tools, and deployment
The Solution
klaw brings Kubernetes-style orchestration to AI agents with a single binary that runs anywhere.| Kubernetes | klaw | Purpose |
|---|---|---|
| Pod | Agent | Unit of deployment |
| Deployment | AgentBinding | Desired state specification |
| Service | Channel | How agents receive work |
| Node | Node | Where agents run |
| Namespace | Namespace | Isolation boundary |
| kubectl | klaw | CLI interface |
Key Features
300+ LLM Models
300+ LLM Models
Access Claude, GPT-4, Gemini, Llama, and 300+ more models through a single API using the each::labs router. No need to manage multiple API keys or endpoints.
Multi-Channel Support
Multi-Channel Support
Deploy agents to CLI, Slack, Discord, Telegram, or REST API. Same agent, multiple surfaces.
Built-in Tools
Built-in Tools
Agents come with powerful capabilities out of the box: bash execution, file operations, web search, and more.
Skills System
Skills System
Composable capability bundles that can be installed and combined to create specialized agents.
Distributed Mode
Distributed Mode
Scale from a single machine to enterprise clusters with controller-node architecture.
Single Binary
Single Binary
One ~20MB binary with zero external dependencies. No Python, no Node.js, no containers required.
