Agent Skills: Testing Skill

Testing Skill

Comprehensive testing domain for pytest-based projects. Provides test generation, review, execution, and specialized testing workflows with automatic routing based on user intent.

Auto-Activation Keywords

This skill activates automatically when you mention:

• test, tests, testing
• coverage, test coverage
• pytest, unittest
• integration test, unit test, e2e test
• performance test, benchmark, load test
• security test, penetration test
• test review, test quality

Routing Logic

Based on user intent, this skill routes to appropriate workflows:

Test Generation

Keywords: "generate tests", "create tests", "write tests", "test coverage"

• Simple test generation → Use /testing/generate workflow
• Complex test strategy → Invoke test-engineer agent

Test Review

Keywords: "review tests", "test quality", "coverage gaps", "improve tests"

• Review existing tests → Use /testing/review workflow
• Comprehensive quality audit → Invoke test-engineer agent

End-to-End Testing

Keywords: "e2e test", "end-to-end", "full workflow", "integration test", "pipeline test"

• E2E test generation → Use /testing/e2e workflow
• Complex multi-component → Invoke test-engineer agent

Performance Testing

Keywords: "performance test", "benchmark", "load test", "memory usage", "speed test"

• Performance test creation → Use /testing/performance workflow
• Comprehensive performance analysis → Invoke test-engineer agent

Security Testing

Keywords: "security test", "vulnerability", "penetration test", "OWASP", "injection"

• Security test generation → Use /testing/security workflow
• Full security audit → Invoke security-auditor agent

Test Execution

Keywords: "run tests", "execute tests", "test results", "pytest command"

• Quick reference → See context/pytest-commands.md
• Custom test execution → Use bash with pytest commands

Workflow Quick Reference

# Generate comprehensive tests
/testing/generate [module-path]

# Review test quality
/testing/review [test-directory]

# Create E2E tests
/testing/e2e [workflow-name]

# Create performance tests
/testing/performance [component]

# Create security tests
/testing/security [attack-vector]

Complex Task Delegation

For complex multi-step testing tasks, invoke the test-engineer agent via Task tool:

Use test-engineer agent when:
- Designing comprehensive test strategy across multiple components
- Creating complete test suites from scratch (unit + integration + e2e)
- Optimizing test performance and coverage simultaneously
- Integrating multiple testing types (functional + performance + security)

Supporting Context

• Pytest patterns: See context/pytest-patterns.md for common test patterns
• Common commands: See context/pytest-commands.md for frequently used commands
• Coverage standards: Minimum 80% coverage required (see /standards/python.md)

Integration Points

Agents

• test-engineer: Comprehensive test strategy and generation
• security-auditor: Security-focused testing and vulnerability analysis

MCP Tools

• mcp__zen-core__chat: Multi-model consensus for complex test design decisions

Hooks

• TDD enforcement: Pre-Write/Edit hook enforces test-first development

Standards

• Python testing standards: /standards/python.md (80%+ coverage requirement)
• Linting standards: /standards/linting.md (pytest configuration)

Testing Standards Summary

Required for all projects:

• Minimum 80% test coverage
• Unit tests in tests/unit/ (< 1s each)
• Integration tests in tests/integration/ (< 5s each)
• E2E tests in tests/e2e/ (marked @pytest.mark.e2e)
• Proper pytest markers (@pytest.mark.unit, .integration, .slow, .security, .perf)
• AAA pattern (Arrange-Act-Assert)
• Clear test naming (test_should_[expected]when[condition])

Fixture management:

• Use existing fixtures from conftest.py
• No hardcoded test data
• Appropriate fixture scope (function, class, module, session)

Test execution:

# Standard test run with coverage
poetry run pytest -v --cov=src --cov-report=html --cov-report=term-missing --cov-fail-under=80

# Fast dev cycle (skip slow tests)
poetry run pytest -m "not slow"

# Specific test categories
poetry run pytest tests/unit/          # Unit tests only
poetry run pytest tests/integration/   # Integration tests
poetry run pytest -m security          # Security tests
poetry run pytest -m perf              # Performance tests

Examples

Example 1: Generate tests for new module

User: "Generate comprehensive tests for the payment processor module"
→ Skill routes to /testing/generate workflow
→ Workflow creates unit, integration, and security tests
→ Ensures 80%+ coverage for payment processor

Example 2: Review test quality

User: "Review the tests for the authentication system and suggest improvements"
→ Skill routes to /testing/review workflow
→ Workflow analyzes coverage, patterns, fixture usage
→ Provides actionable recommendations

Example 3: Complex test strategy

User: "Design a comprehensive testing strategy for the new RAG pipeline"
→ Skill invokes test-engineer agent
→ Agent creates multi-tier testing approach:
  - Unit tests for embeddings, chunking, retrieval
  - Integration tests for pipeline flows
  - E2E tests for complete query-response cycles
  - Performance tests for large document handling

This skill consolidates test-generator, test-reviewer, e2e-tester, performance-tester, and security-tester into a unified testing domain with intelligent routing.