CLI-7: Add evals.json files to skills for quality assurance and testing

[szmania]

Summary

This PR adds evals.json files to skills for quality assurance and testing, following the pattern established in the marketingskills repository.

Changes Made

1. Added evals directory structure

• Created evals/ directory in cecli/skills/web/playwright-cli/
• Added evals.json file with sample test cases

2. evals.json structure

The evals.json file follows the standard format with:

• skill_name: Identifies which skill is being evaluated
• evals: Array of test cases, each containing:
  • id: Unique identifier for the test
  • prompt: The user query to test
  • expected_output: Description of expected behavior
  • assertions: List of specific checks to verify
  • files: Expected output files

3. Sample test case added

• Test case for opening playwright documentation and taking a screenshot
• Demonstrates the expected format for skill evaluations

Problem Solved

Before this change, cecli skills lacked structured evaluation tests to verify their effectiveness and consistency. This PR establishes the foundation for:

• Quality assurance for skill performance
• Regression testing when skills are modified
• Clear documentation of expected behavior
• Automated validation of skill outputs

Testing

The evals.json structure has been validated against the marketingskills repository pattern.

Related Issue

Implements CLI-7: Add evals.json files to all skills for quality assurance and testing.