Playwright Verifier

playwright-verifier Agent

Type: Hidden helper sub-agent (not user-invokable)
Platforms: Copilot (.github/agents/playwright-verifier.agent.md), Claude Code (.claude/agents/playwright-verifier.md)
Invoked by: web-issue-fixer after applying each fix

Purpose

Closed-loop fix verification using Playwright. After a fix is applied, navigates to the fixed element, runs a targeted axe-core assertion or behavioral check, and reports PASS/FAIL/REGRESSION. Replaces screenshot-based verification with deterministic, assertable browser tests.

Verification Flow

1. Receive fix context: rule ID, CSS selector, dev server URL, fix type
2. Run the appropriate verification tool based on fix type
3. Compare pre-fix and post-fix state
4. Return verdict: PASS, FAIL, or REGRESSION

Fix Type → Verification Tool Mapping

Fix Type	Tool Used	What Is Checked
contrast	run_playwright_contrast_scan	Computed colors, contrast ratio meets threshold
keyboard	run_playwright_keyboard_scan	Element in tab order, no traps introduced
aria	run_playwright_a11y_tree	Role, name, state in accessibility tree
structure	run_playwright_a11y_tree	Heading hierarchy, landmark structure
state	run_playwright_state_scan	Dynamic content accessible after interaction
viewport	run_playwright_viewport_scan	Reflow and touch targets at all widths

Test Code Generation

After a verified PASS, generates a Playwright test file that encodes the assertion for regression prevention. Test files are saved to tests/a11y/ for inclusion in the project's test suite.

Dependencies

Same as playwright-scanner — requires playwright, optionally enhanced by @axe-core/playwright.

Related Agents

• web-issue-fixer — Orchestrator that dispatches this agent
• playwright-scanner — Scanning counterpart