# unotest documentation

> AI-native E2E testing for web and iOS. Your AI agent writes the tests — you review and commit. Plain .js scenarios in your repo, a local viewer, an MCP server, full human control.

Each link below points to the raw Markdown of a page. For the entire docs in one file, fetch https://docs.unotest.com/llms-full.txt

## Start here
- [How it works](https://docs.unotest.com/start/how-it-works.md): The architecture behind unotest — MCP, semantic perception, a sandboxed engine, and local-first execution.
- [Install & setup](https://docs.unotest.com/start/install.md): Requirements, browser choices, and wiring your editor to unotest over MCP.
- [Introduction](https://docs.unotest.com/start/introduction.md): What unotest is — AI-native E2E testing where your agent writes the tests and you stay in control.
- [Quick start — iOS](https://docs.unotest.com/start/quickstart-ios.md): From a built .app to a passing E2E test on the iOS Simulator, written by your agent.
- [Quick start — web](https://docs.unotest.com/start/quickstart-web.md): From zero to a passing E2E test in your web app, written by your agent.

## Concepts
- [Collections](https://docs.unotest.com/concepts/collections.md): Group scenarios and run them as a set — smoke, regress, and more.
- [Debugging & self-healing](https://docs.unotest.com/concepts/debugging.md): One pause-on-failure debugger for both the agent and you — and why fixes are never silent.
- [Failure bundles](https://docs.unotest.com/concepts/failure-bundles.md): The evidence captured when a run fails — screenshots, console, semantic DOM, and trace.
- [Helpers — flows & mocks](https://docs.unotest.com/concepts/helpers.md): Reusable JavaScript functions for repeated journeys and for seeding data.
- [Scenarios & step()](https://docs.unotest.com/concepts/scenarios.md): How a unotest scenario is structured — readable intent on top, precise DSL underneath.
- [Stable selectors](https://docs.unotest.com/concepts/selectors.md): The selector priority that keeps tests resilient, and the linter that enforces it.
- [Variables & secrets](https://docs.unotest.com/concepts/variables.md): Keep credentials out of code and run the same test across environments.

## Guides
- [Auth & cached login](https://docs.unotest.com/guides/auth.md): Reuse a logged-in session instead of signing in at the start of every test.
- [Write a test by hand](https://docs.unotest.com/guides/by-hand.md): Author a scenario yourself — the format is just readable JavaScript.
- [Run in CI](https://docs.unotest.com/guides/ci.md): Run unotest scenarios and collections in continuous integration.
- [Run smoke / regress](https://docs.unotest.com/guides/collections.md): Group scenarios into collections and run them as a set.
- [Debug a failing test](https://docs.unotest.com/guides/debug.md): Use the pause-on-failure debugger — in the viewer or with your agent.
- [Write your first test](https://docs.unotest.com/guides/first-test.md): Let your agent write and verify a real E2E test, step by step.
- [Reuse flows & seed data](https://docs.unotest.com/guides/flows-and-data.md): Factor repeated journeys into flows and put the app into a known state with mocks.
- [Multi-tab & iframes](https://docs.unotest.com/guides/multi-context.md): Drive multiple tabs and work inside nested iframes.

## The Viewer
- [Step debugger](https://docs.unotest.com/viewer/debugger.md): Gutter breakpoints, pause-on-failure, and live inspection.
- [Inspector & artifacts](https://docs.unotest.com/viewer/inspector.md): The failure evidence shown per run — screenshot, semantic DOM, console, trace.
- [Overview & launch](https://docs.unotest.com/viewer/overview.md): A local IDE-style viewer for your tests — no cloud, no account.
- [Running tests](https://docs.unotest.com/viewer/running.md): Run from the UI and watch results stream live, step by step.
- [Shortcuts & themes](https://docs.unotest.com/viewer/shortcuts.md): Keyboard shortcuts and theme switching in the viewer.
- [Terminal (AI inside)](https://docs.unotest.com/viewer/terminal.md): A docked terminal with local, Claude and Codex sessions — drive the agent without leaving the viewer.
- [Tour](https://docs.unotest.com/viewer/tour.md): The activity bar, document tabs, and inspector panel.

## Reference
- [CLI — mobile](https://docs.unotest.com/reference/cli-mobile.md): Every npx @unotest/mobile subcommand (iOS).
- [CLI — web](https://docs.unotest.com/reference/cli-web.md): Every npx @unotest/web subcommand, with flags and examples.
- [Configuration](https://docs.unotest.com/reference/config.md): Every field in unotest.config.{js,mjs,ts}.
- [DSL reference](https://docs.unotest.com/reference/dsl.md): Every function available inside a scenario, grouped — the unotest web vocabulary.
- [Linter rules](https://docs.unotest.com/reference/linter.md): The scenario linter rules and their default severities.
- [MCP tools](https://docs.unotest.com/reference/mcp-tools.md): Every MCP tool the agent uses, grouped.

## For your AI agent
- [Authoring guide](https://docs.unotest.com/agent/authoring.md): The rules an agent should follow when writing unotest scenarios.
- [Connect your editor](https://docs.unotest.com/agent/connect.md): Wire Claude Code, Cursor or Codex to the unotest MCP server.
- [Machine-readable docs](https://docs.unotest.com/agent/machine.md): Every page as Markdown, plus llms.txt and llms-full.txt for agents.
- [Overview](https://docs.unotest.com/agent/overview.md): How an AI agent uses unotest — MCP-native, with a human in the loop.
- [Worked example — authoring over MCP](https://docs.unotest.com/agent/worked-example.md): The exact MCP tool-call sequence to explore an app, record a scenario, save it, and verify it runs green.

## Resources
- [Changelog](https://docs.unotest.com/resources/changelog.md): Notable changes to unotest.
- [Ecosystem & packages](https://docs.unotest.com/resources/ecosystem.md): The packages that make up unotest and how they fit together.
- [Troubleshooting](https://docs.unotest.com/resources/troubleshooting.md): Common issues and known limitations.