The console
L80 tui is a full-screen console for everything L80 does. Each screen runs the same code as its command, so anything you do here you can script later.
Free, no token
Open it
$ L80 tuiLinux and macOS. Without a terminal it refuses rather than guessing; L80 with no arguments still prints the command list. NO_COLOR turns colour off.
Keys
| key | does |
|---|---|
| ↑ ↓ | Move through the sidebar; each screen previews live in the pane |
| enter | Open the screen |
| esc | Back |
| 1–8 | Jump to a screen |
| ? | Help |
Screens
| screen | what you can do |
|---|---|
| Home | Server, token and tool status at a glance, plus your last publishes. |
| Benchmark | Fill in a form, watch the run stream, then publish or dry-run. |
| Vast | Pick a spec, see cost caps before renting, follow runs, destroy boxes. |
| pwdet | Scan a model, read the verdict and findings, publish the result. |
| Dashboard | Attach to the telemetry view or start the sampler. |
| History | Every page you published, with its trust tier. v requests verification. |
| Verify | Other keys' requests: p reproduces; verifiers stamp with a or reject with x. |
| Tools | How each tool resolves, and prefetch it. |
Guides for each: Benchmark, Vast, pwdet, Dashboard, Verify, Tools.
What it asks first
- Renting, destroying and publishing always ask, with the safe choice focused and the cost caps or the privacy note in front of you.
- Tool output streams into a scrollable log and is sanitised: a subprocess cannot move the cursor or retitle the window.
vast sshand the dashboard take over the terminal and hand it back when they exit.