Expect

Let agents test your code in a real browser.

One command scans your unstaged changes or branch diff, generates a test plan, and runs it against a live browser.

See it in action →

Install

npx -y expect-cli@latest init

Usage

Usage: expect-cli [options] [command]

Options:
  -m, --message <instruction>   natural language instruction for what to test
  -f, --flow <slug>             reuse a saved flow by slug
  -y, --yes                     skip plan review, run immediately
  -a, --agent <provider>        agent provider to use (claude or codex)
  -t, --target <target>         what to test: unstaged, branch, or changes (default: changes)
  --verbose                     enable verbose logging
  -v, --version                 print version
  -h, --help                    display help

Commands:
  init                          install expect globally and set up skill

Examples:
  $ expect-cli                                          open interactive TUI
  $ expect-cli -m "test the login flow" -y              plan and run immediately
  $ expect-cli --target branch                          test all branch changes
  $ expect-cli --target unstaged                        test unstaged changes

Set `NO_TELEMTRY=1` to disable analytics events.

How it works

 ┌─────────────┐     ┌──────────────┐     ┌──────────────┐     ┌─────────┐
 │ Scan changes│────▶│ Generate plan│────▶│ Run in browser│────▶│ Report  │
 │ (git diff)  │     │ (AI agent)   │     │ (Playwright)  │     │ (pass/fail)│
 └─────────────┘     └──────────────┘     └──────────────┘     └─────────┘

Expect reads your unstaged changes or branch diff, sends them to an AI agent (Claude Code or Codex CLI), and generates a step-by-step test plan describing how to validate the changes. You review and approve the plan in an interactive TUI, then the agent executes each step against a live browser - using your real login sessions so there's no manual auth setup. Every session is recorded so you can replay exactly what happened.

Pass -y to skip plan review and run headlessly in CI. Exits 0 on success, 1 on failure.

Resources & Contributing Back

Want to try it out? Check out our demo.

Find a bug? Head over to our issue tracker and we'll do our best to help. We love pull requests, too!

We expect all contributors to abide by the terms of our Code of Conduct.

→ Start contributing on GitHub

License

FSL-1.1-MIT © Million Software, Inc.