Klaw.Sh

Klaw is an agent orchestration tool that manages AI agents with kubectl style commands. It is distributed as a single binary of about 20 MB with no Python or Docker dependency, installed with a shell script, and the vendor labels it as in public beta. The command set follows Kubernetes conventions, with get, describe, logs and apply, so agents can be listed, inspected and scheduled from a terminal. Agents can also be controlled from Slack by mentioning @klaw, through a TUI and through a REST API. Namespaces isolate teams and secrets, and Podman containers can sandbox the filesystem. A built in cron scheduler runs agents hourly, daily or weekly, and a controller and node architecture supports distributed mode, from a laptop to a cluster. The vendor lists more than 300 LLM models, including Claude, GPT, Gemini and Llama, through the each::labs Router or direct provider access. Example use cases on the site include CRM lead scoring, scheduled revenue reports from Stripe and competitor pricing monitoring. It is aimed at engineering and operations teams that already work with kubectl and want the same workflow for running agents in production. The site links to GitHub and documentation. No pricing page is published, so pricing is unpublished.