Interactive mode (coming soon)
A terminal UI for browsing evaluations and dispatching autousers — planned for a future release.
Not shipped yet
autousers tui is on the roadmap but not part of the current @autousers/cli release. Use the non-TUI commands documented in the command reference for now — every workflow the TUI will expose is already available as a flag-driven subcommand.
What is planned
- Full-screen browser for evaluations, autousers, and templates
- Live dashboard for in-flight autouser runs with token/cost telemetry
- Inline cost preview before any --commit, mirroring the MCP dryRun-first rule
- Keyboard-first navigation with search, filter, and quick actions
Use the commands today
Until the TUI ships, the core surface is the same set of subcommands the TUI will sit on top of:
autousers eval list
autousers eval get <id>
autousers eval create
autousers eval run <id> # dry-run preview
autousers eval run <id> --commit # confirm + queue
autousers eval results <id>
autousers usageThe non-TUI surface covers every TUI workflow we have planned
Tell us what you want
We are designing the TUI from real user workflows, not assumptions. If there is a specific view or shortcut you would want — bulk dispatch, calibration drill-downs, share-management UI, anything — please send it over at https://autousers.ai/contact and we will fold it into the design.
Prefer not to wait? Connect Autousers to Claude or Cursor over MCP and drive the same workflows from natural language today — see the MCP overview.
Was this article helpful?