Test Analyze

What it does

This skill, Test Analyze, helps AI agents analyze test results from various testing frameworks and create a systematic plan for fixing identified issues. It parses test result files, categorizes problems by type and severity, and uses a planning tool (Zen planner via MCP) to break down fixes into actionable steps. Finally, it delegates these fix tasks to specialized subagents based on the nature of the issue.

When to use it

• After running unit tests to identify and address code failures.
• Following integration testing to pinpoint issues between different system components.
• When conducting accessibility audits using Playwright or similar tools, ensuring WCAG compliance.
• To analyze security scan results (e.g., OWASP ZAP) and prioritize remediation efforts.
• After performance tests (Lighthouse, k6) to identify bottlenecks and areas for optimization.

Key capabilities

• Test Result Parsing: Supports XML, JSON, HTML, and text formats.
• Issue Categorization: Classifies issues by type (accessibility, unit, integration, e2e, security, performance) and severity.
• Fix Planning: Leverages a planning tool via MCP to create structured fix strategies.
• Subagent Delegation: Automatically routes tasks to specialized subagents like code-review, security-audit, system-debugging, and documentation agents.
• Test Type Detection: Can automatically detect the type of tests being analyzed if not explicitly specified.

Example prompts

• /test:analyze ./test-results/ --type accessibility
• /test:analyze ./coverage/junit.xml --type unit --focus authentication
• /test:analyze ./test-output/

Tips & gotchas

• The results-path parameter is required to specify the location of test results.
• While the skill attempts to auto-detect the test type, providing it explicitly with --type can improve accuracy and ensure correct subagent delegation.