What Is a Test in This Platform?

In Rihario, a "test" is actually an AI-powered exploration, not a scripted test case. The AI explores your app, interacts with elements, and reports what it finds. There's no pass/fail - just findings, evidence, and your judgment.

Traditional Test vs Rihario Exploration

Traditional Test (Playwright, Cypress, etc.)

// Traditional test script
test('login works', async () => {
await page.goto('/login')
await page.fill('#email', 'user@example.com')
await page.fill('#password', 'password123')
await page.click('#submit')
await expect(page.locator('#dashboard')).toBeVisible()
// Result: PASS or FAIL
})

Result: PASS or FAIL. Deterministic. Same input always produces same output.

Rihario Exploration

Rihario exploration:

URL: https://app.com/login

Instructions: "check if login works"

→ AI explores the page → Finds email field (automatically) → Types test email → Finds password field → Types test password → Clicks submit button → Observes what happens → Reports findings

Result: "Explored login page. Form submitted successfully. Redirected to dashboard. No issues found."

Result: Findings report. Probabilistic. Same input produces similar results, but not identical.

Key Differences

1. No Pass/Fail Status

Rihario doesn't say "PASS" or "FAIL". Instead, it says:

"I explored 5 pages and found 2 issues"
"Everything looks good - no problems detected"
"Hit a blocker - couldn't proceed past CAPTCHA"

You decide what matters. The AI reports findings, you interpret them.

2. No Scripts or Assertions

Traditional tests require you to write:

Scripts defining exact steps
Assertions checking specific conditions
Selectors targeting exact elements

Rihario doesn't use scripts. You provide:

A URL to explore
Optional instructions in plain English
That's it

3. AI Decides What to Do

The AI analyzes the page and decides:

What elements are interactive
What looks important vs decorative
What to click, type, or check
When something looks wrong

You're not dictating actions - you're setting a goal and letting the AI figure out how to achieve it.

4. Adapts to Changes

If your app changes (new button text, different layout, updated selectors), the AI adapts automatically. Traditional tests break and need updates. Rihario explorations continue working.

5. Probabilistic, Not Deterministic

Run the same exploration twice:

Traditional test: Exact same steps, exact same result every time
Rihario: Similar steps, similar findings, but not identical

The AI might explore in slightly different orders, take different paths, or notice different things. This is by design - it's exploring, not executing a script.

What Actually Happens During a "Test"

You start an exploration - Provide URL and optional instructions
AI analyzes the page - Understands structure, finds interactive elements
AI explores - Clicks, types, navigates based on what it finds
AI detects issues - Notices errors, broken layouts, problems
AI collects evidence - Screenshots, logs, DOM snapshots
You see results - Findings report with evidence
You decide - Fix issues, ignore false positives, or explore more

Terminology We Use

Even though we call them "tests" for simplicity, here's what we really mean:

"Test" = Exploration session
"Test Run" = Single exploration execution
"Step" = Individual action the AI took
"Issue" = Problem the AI detected
"Finding" = Something the AI noticed (might be an issue, might not)
"Evidence" = Screenshots, logs, DOM snapshots supporting a finding

Why This Matters

Understanding that Rihario uses explorations, not scripted tests, helps set proper expectations:

You won't get guaranteed coverage - The AI explores what it finds, not every possible scenario
Results require human judgment - Not everything the AI flags is a real problem
It's probabilistic - Same exploration might produce slightly different results
It's exploratory - Focused on finding issues, not verifying specific behaviors

Next Steps