Kranz: a CLI for managing local dev services
I posted Kranz here as a TUI for local development services and logs. Since then, I've added CLI commands and MCP access to the running services.
The setup I had in mind when I started Kranz is now working. You can start services in the dashboard, detach and manage them from the shell, or give a coding agent access through MCP. All three use the same running services.
The GIF shows this with a small demo project:
kranz status
kranz plan api --operation restart
kranz restart api
kranz runs api --limit 2
kranz logs api --run -1 --tail 3
kranz actions run api/test --param suite=smoke
Here, restarting api also restarts web and worker. plan shows that before you do it. Each service run has separate saved logs, so you can check what happened before a restart. --run -1 selects the latest run.
The last command runs a test action configured in this demo. It accepts a suite parameter, which Kranz checks against the allowed values.
At the end of the GIF, an MCP client reads the service status, API logs, and test result. A coding agent can access those too. These are just a few of the CLI commands. You can also check configs, inspect dependency graphs and ports, get JSON output for scripts, and more.
Recorded on v0.16.4 with demo services.
If you try it on your dev stack, I'd like to know which commands or output feel awkward. If it looks useful to you, I'd also appreciate a star on GitHub.
As in the first post: coding agents wrote most of the Go code. I designed the behavior and UI and tested the result.