From ralph-playwright
Analyze a journey trace and its screenshots to produce a signal report. Use when you have a journey trace from a previous execute run and want to analyze it separately. Reads screenshots and accessibility snapshots to identify anomalies, regressions, a11y violations, and UX issues.
How this skill is triggered — by the user, by Claude, or both
Slash command
/ralph-playwright:reflectThis skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
Path to a journey trace YAML file (from a previous execute run):
Path to a journey trace YAML file (from a previous execute run):
.playwright-cli/2026-03-21-explore-checkout-flow/journey-trace.yamlRead the journey trace YAML. Verify it conforms to the journey-trace schema (has id, timestamp, steps, summary).
For each step in the trace:
screenshot path) — look for visual anomalies, layout issues, error states.md file at the snapshot path) — check element structure, labels, roles, ARIA attributesFor each finding, classify as:
| Type | When |
|---|---|
anomaly | Unexpected behavior, visual glitches, broken layouts |
regression | Something that previously worked now fails (requires baseline comparison) |
a11y_violation | WCAG non-compliance: missing labels, broken tab order, contrast |
ux_issue | Confusing navigation, dead ends, unclear feedback |
error | Console errors, failed steps, broken interactions |
Assign severity:
critical: Blocks core functionality or causes data losshigh: Major usability or accessibility barriermedium: Noticeable issue but workaround existslow: Minor cosmetic or best-practice issueWrite to .playwright-cli/<session>/signal-report.yaml following the signal-report schema:
trace_id: "<from trace>"
timestamp: "<now ISO-8601>"
signals:
- type: <type>
severity: <severity>
title: "<short title>"
description: "<detailed description>"
evidence:
steps: [<step indices>]
screenshots: ["<screenshot filenames>"]
tags: [<relevant tags>]
summary:
total_signals: <N>
by_severity: { critical: N, high: N, medium: N, low: N }
recommendation: "<actionable recommendation>"
== Signal Report for <session> ==
Trace: <trace_id> | Steps: <N> | Duration: <ms>
Signals: N total
🔴 Critical: N
🟠 High: N
🟡 Medium: N
⚪ Low: N
<signal details>
Recommendation: <recommendation>
Next: Use /ralph-playwright:capture to promote screenshots, or pipe this
report to the act primitive for automated issue creation.
npx claudepluginhub cdubiel08/ralph-hero --plugin ralph-playwrightAudits application workflows for bugs, data safety, performance, and round-trip completeness. Discovers workflows, traces user journeys, and rolls up cross-cutting patterns.
Debugs and fixes flaky Playwright E2E tests using LLM reports from GitHub Actions and Datadog. Use for investigating intermittent failures, triaging flakiness, or stabilizing tests.
Drives a real Chrome session against a running web app to find bugs, UX issues, a11y problems, and perf regressions. Outputs structured findings JSON for downstream triage.