Where raw Playwright is the right choice
Raw Playwright gives teams complete control. Every aspect of how tests are authored, structured, and run is in the team's hands. For teams with dedicated QA engineers who are comfortable maintaining a significant volume of browser automation code, that control is a genuine asset.
- Complete flexibility in test structure, helpers, and execution patterns
- No abstraction layer between the test author and the browser
- Well-suited to teams with dedicated automation engineers
- Strong fit when tests require complex custom logic that generic authoring cannot express
- Extensive ecosystem of plugins, reporters, and integrations
Where raw Playwright gets expensive
The cost of raw Playwright is proportional to the volume of tests and the rate of UI change. At small scale with a stable UI, the overhead is manageable. As the product grows and the UI evolves, the maintenance burden compounds quickly.
Every UI change that touches a selector requires a manual test update. Every new flow requires a new hand-authored test file. Non-engineering stakeholders cannot easily read or review what the tests are protecting. And when a test fails, an engineer has to read through automation code to understand what user behavior is actually at risk.
- Every new flow requires hand-authored automation code from scratch
- Selector drift creates repetitive, low-value maintenance work after UI changes
- Test files are often too mechanical for meaningful cross-discipline review
- Intent is embedded in code rather than expressed in readable form
- AI assistance produces disposable output unless the team has a strong review workflow
What Assert changes
Assert uses Playwright as its execution engine — you still get real browser automation with all of Playwright's reliability. The difference is in how tests are authored and maintained. Scenarios are defined in plain-English Markdown, kept in the repo, and used to generate the Playwright layer automatically.
That changes the maintenance equation. When the UI changes, engineers update a readable scenario file rather than debugging selector failures. When a new flow needs coverage, it can be drafted in seconds by a developer or an AI agent and reviewed in a pull request like any other change.
Assert does not give you more control over test execution than raw Playwright — it gives you less, deliberately. In exchange, authoring is faster, maintenance is cheaper, and the artifacts engineers review are legible to everyone on the team.
How to decide
Choose raw Playwright if your team has dedicated automation engineers who are comfortable with selector-level authoring, you need complex custom test logic that a higher-level abstraction cannot express, or you are building shared test infrastructure that other tools will build on top of.
Choose Assert if your team is spending too much time maintaining brittle tests, you want developers and AI agents to be able to create and update tests as part of normal development, or you need the test artifacts to be reviewable by engineers, product managers, and QA leads alike.
FAQ
Does Assert replace Playwright?
No. Assert uses Playwright as its execution engine. When Assert runs a test, it generates a Playwright test from the scenario file and executes it using Playwright's browser automation. You get all of Playwright's reliability and browser support; Assert handles the authoring and maintenance layer above it.
Can I use Assert alongside existing Playwright tests?
Yes. Assert does not require you to replace your existing test suite. Many teams start by adding Assert scenarios for new flows while keeping their existing Playwright tests for flows that already have solid coverage.
Is raw Playwright ever the better choice for teams using AI?
If your team has a disciplined workflow for reviewing and committing AI-generated Playwright code, raw Playwright with AI assistance can work well. The risk is that AI-generated Playwright code tends to be opaque and hard to maintain — correct today, confusing in six months. Assert's Markdown layer keeps AI output readable and reviewable regardless of where or how it was generated.
Put the workflow in your repo, not in a chat transcript
Assert is strongest when scenarios become durable project assets: readable Markdown in the repo, generated execution underneath, and result inspection in the dashboard.