Bring your own AI with Level CLI

This release gives teams new ways to extend Level and coordinate work across their environment. Use the Level CLI with supported AI coding agents to manage devices, build automations, and handle repetitive tasks using the AI tools your team already prefers. The new Run automation on another device action also makes it possible to connect workflows across managed devices and pass variables between automations. Rounding out the release are usability and performance improvements, along with fixes that make Level more reliable across day-to-day workflows.

Major feature

Bring your own AI with Level CLI

Level CLI gives you direct terminal access to your devices, alerts, scripts, automations, monitors, updates, groups, and more. Run commands yourself, or connect the CLI to Codex, Claude Code, Gemini, or another coding agent and describe the outcome you need. Your coding agent can inspect device context, investigate alerts, run diagnostics, propose fixes, execute scripts across groups, and wait for the results. It can also create and update monitors and automations as code, with diff and dry-run support so you can review changes before applying them. Because humans and coding agents use the same CLI, actions remain governed by the permissions of the Level user running them and are recorded in Level’s audit history. From troubleshooting a single endpoint to patching or remediating an entire fleet, Level CLI brings the full power of Level into your terminal. Read more about the CLI in our documentation: https://docs.level.io/en/articles/16601525-level-cli-lvl-get-started

Major feature

Run automation on another device

Build workflows that coordinate work across multiple devices. With the new Run automation on another device action, an automation running on one endpoint can launch a selected automation on another managed device, such as a server or dedicated runner. Map variables between the two automations to carry important context forward and keep the entire process connected.