Skip to content

Write tests in plain English. An AI agent runs them.

Describe what to test in a TOML file. Bugatti handles setup, drives an AI agent through each step, and gives you structured OK/WARN/ERROR results with a full audit trail.

Write steps in plain English

Describe what to test in a .test.toml file. No selectors, no page objects, no test framework DSL.

An AI agent runs them

Bugatti drives an AI coding agent through your test plan. The agent figures out how to verify each step.

Get structured results

Every step reports OK, WARN, or ERROR. Full transcripts, logs, and reports saved per run.

Built-in infrastructure

Setup commands, long-lived servers with readiness polling, state checkpoints, composable test files.

login.test.toml
name = "Login flow"
[[steps]]
instruction = "Navigate to /login and verify the page loads"
[[steps]]
instruction = "Enter valid credentials and submit the form"
[[steps]]
instruction = "Verify you are redirected to the dashboard"
Terminal window
bugatti test login.test.toml

Manual QA doesn’t scale. Traditional E2E test frameworks are brittle and expensive to maintain. Bugatti sits in between — you describe what to test in natural language, and an AI agent handles the how.