HTTP tunneling

Open an HTTP tunnel to any web service reachable from the device (on localhost or its private LAN):

# Default: target localhost:80
thinr device http <deviceId>

# Specific port
thinr device http <deviceId> 8080

# Specific address reachable from the device
thinr device http <deviceId> http://192.168.1.45:8080

# An HTTPS target
thinr device http <deviceId> https://localhost:9443

The CLI prints the tunnel URL on your ThinRemote server and (by default) opens it in your browser. Add --no-open to suppress the browser launch; -p <port> to pin the server-side port, which keeps the URL stable across runs.

The tunnel lives until you press Ctrl+C; while it's up, whatever the target exposes is protected by its own login. The plumbing is the agent's tunnels feature: the device dials the target, nothing listens on it.

Typical targets

Anything with a web UI that was never meant to be on the internet:

• Dashboards: Grafana (3000), Node-RED (1880), Portainer (9000).
• Embedded panels: a router's admin page, a PLC's web server, a camera, all addressed by their LAN IP next to the device.
• Your own app during development or debugging.

thinr device http edge-gw-17 3000                      # Grafana on the device
thinr device http edge-gw-17 http://192.168.1.1        # the site router's admin UI

Multiple tunnels

Each tunnel is independent: run several commands in parallel (one per terminal, or backgrounded in scripts) and each gets its own relay port.

Not HTTP?

Databases, MQTT, custom protocols: use TCP/TLS tunneling. SSH specifically has its own workflow.