How to use the Teltonika RMS device manager agent skill

What is an agent skill and what does this one do?

An agent skill is a folder containing a SKILL.md file that gives an AI agent instructions for a specific tool. This skill covers the Teltonika RMS (Remote Management System) cloud service, which manages industrial routers, gateways and IoT devices.

Once installed, the agent can list devices, check their status and location, review alerts and logs, and manage users, companies, tags, SMTP settings, remote access sessions and reports. It does this by running cli-anything-rms commands against the RMS REST API. Many of these commands create, update or delete live records.

• Devices: list, get details, update name or tag, delete
• Alerts and logs: list, get, and manage alert configurations
• Location: current device location and location history
• Remote access, hotspots, configs, reports, files, SMTP and credits
• Auth and config: test connectivity and manage the API token

Install the Teltonika RMS skill

Download the skill ZIP from this page and unzip it into one of two locations. Use ~/.claude/skills/ to make the skill available across your projects, or use a project's .claude/skills/ folder to limit it to that project.

Alternatively, upload the ZIP in the Claude apps under Settings → Capabilities → Skills.

The skill also needs the command-line package. The source installs it with pip from GitHub. Python with pip is required.

• Option 1: unzip into ~/.claude/skills/ for personal use
• Option 2: unzip into a project's .claude/skills/ folder
• Option 3: upload the ZIP under Settings → Capabilities → Skills in the Claude apps

Set up authentication with your API token

The tool needs your RMS API token. You can provide it in one of two ways. Setting the RMS_API_TOKEN environment variable works, and so does saving it in the tool's configuration with the config set command.

Treat the token carefully. It grants account-level access, so do not paste it into shared agent sessions or into logs.

• Environment variable: set RMS_API_TOKEN
• Configuration: store it with config set api_token
• Check where config is kept with config path

Check the connection before you ask for changes

Run the connectivity and auth commands first. They confirm the token works and that the RMS REST API is reachable from your machine. If these fail, no other command will succeed because every request requires authentication.

Once the checks pass, start with read-only requests such as listing devices or checking alerts.

Ask for what you need in plain language

You do not need to memorise the command list. Describe the outcome you want, such as checking which routers are offline or pulling recent alerts for one device, and the agent will choose the matching cli-anything-rms command. Name the device ID or tag where you can, because the commands filter on device IDs, tags and status.

For direct use, the tool also supports a JSON output flag and an interactive mode. Read-only listing and get commands are the safest starting point.

• Show all online devices
• Get the details of device 12345 as JSON
• Check alerts for device 12345
• Show the location history for a device, limited to 20 entries

Review changes before the agent runs them

Commands that change data are where mistakes cost you. Devices delete, users delete, passwords update and remote-access create all change live systems, and can cause outages or lockouts. Review each proposed write command before you approve it.

For password changes, the source offers a stdin option that reads the password from standard input, which it describes as safer than passing it on the command line.

• Read before you write: confirm the target ID with a get or list command first
• Be cautious with delete commands on devices, users, tags and companies
• Use the stdin option for passwords rather than typing them into a command
• Do not run this against production fleets without review