Overview

Molar is an autonomous QA and universal browser-automation platform for web apps. Point it at your app and it maps your user flows, writes and runs end-to-end tests, gates pull requests, watches production, and replays failures step by step. Point it at any website and it will also just do the task you describe — fill a form, extract data, log in, download a file — and hand you a live stream and a structured result.

Four products, one account

Product	What you get
Cartographer	Map flows from a URL → verified scenarios → Playwright export
Guard	PR gating, production monitoring, shadow-prod diff, Mender auto-fix
Clones	Safe local stand-ins for Stripe, Twilio, email, auth, and S3
Trace	Step-by-step replay when a run fails

Your account also includes the dashboard, the molar CLI, and MCP for coding agents — these are account tools, not separate products.

Beyond QA — the universal platform

The same VLM-first agent that maps and tests your app is a general-purpose browser operator. These capabilities work on any site, for humans and agents alike:

Capability	What it does	Docs
Universal automation	"Do X on site Y" — run tasks, extract data, validate state, log in, download	Universal automation
Live view & recording	Watch any run in real time; replay with video, network, console, AI decisions	Live view
Migration	Convert existing Playwright/Cypress/Selenium suites to `.molar.md`	Migrate
Agent-native	MCP + CLI with `--json` everywhere so coding agents verify their own work	For agents

Who it's for

Engineering teams shipping daily who need regression coverage without a dedicated SDET.
Coding agents (Cursor, Claude Code, Codex) that write code and need to verify it works in a real browser — the agent verification loop.
Anyone who needs to automate a browser task reliably, with a live stream and a replayable trace.