Add skill-quality-analyzer and skill-security-analyzer to marketplace

[eovidiu]

Summary

This PR adds two new meta skills to the example-skills collection in the marketplace:

1. skill-quality-analyzer

Comprehensive quality analysis tool for Claude Skills that evaluates across five dimensions:

• Structure & Documentation (20%) - SKILL.md quality, examples, resources
• Security (30%) - Uses skill-security-analyzer for automated vulnerability detection
• User Experience (20%) - Clarity, trigger phrases, workflow documentation
• Code Quality (15%) - Organization, maintainability, scripts
• Integration & Tools (15%) - Tool usage, MCP integration, efficiency

Features:

• Three output modes: Comprehensive Report, Interactive Review, Pass/Fail Certification
• Balanced scoring system with letter grades (A-F)
• Actionable recommendations with impact/effort analysis
• Integration with skill-security-analyzer for automated security checks

2. skill-security-analyzer

Security scanner for Claude Code skills detecting 40+ malicious patterns:

• Command injection (os.system, eval, exec)
• Indirect execution (getattr, import)
• Advanced encoding (ROT13, zlib, XOR, AST)
• YAML injection (!!python/object/apply)
• Shell injection (bash -c, python -c)
• Typosquatting detection (Levenshtein distance)
• Time bombs, environment manipulation, sandbox escapes
• Credential theft, data exfiltration

Features:

• 100% detection rate on test suite of 11 malicious samples
• Risk ratings: CRITICAL/HIGH/MEDIUM/LOW/SAFE
• JSON output format for CI/CD integration
• Comprehensive test suite with malicious samples
• Pluggable signature database

Changes

• ✅ Added both skills to .claude-plugin/marketplace.json in example-skills plugin
• ✅ Updated README.md Meta Skills section with descriptions
• ✅ Included all skill files, references, scripts, and test samples (25 files total)

Testing

Both skills have been tested locally:

• skill-quality-analyzer: Evaluated multiple skills across all five dimensions
• skill-security-analyzer: Achieved 100% detection rate on 11 malicious test samples

Use Cases

skill-quality-analyzer:

• Evaluating skills before installation
• Reviewing skills for best practices adherence
• Certifying skills meet quality standards
• Comparing multiple skills for quality

skill-security-analyzer:

• Analyzing skills for security vulnerabilities before installation
• Auditing existing skills for malicious code
• CI/CD integration for automated security checks
• Educational purposes for understanding skill security

---

These meta skills complement the existing skill-creator by providing quality assurance and security validation for the Claude Skills ecosystem.

[tenequm]

@eovidiu why not just add the references from skill-quality-analyzer to the existing skill-creator skill to make it able to both create proper aligned with best practices skills and to verify existing skills too?

[eovidiu]

@eovidiu why not just add the references from skill-quality-analyzer to the existing skill-creator skill to make it able to both create proper aligned with best practices skills and to verify existing skills too?

skill-analyzer does a comprehensive report or an run interactively to analyze a skill just created or one from an unknown market. Doens't really fit being packed with the skill-creator

[kele527]

在正文部分不应该加这个：When to Use This Skill 吧，因为能看到正文说明skill已经被使用了，这个时候再告诉他何时使用此技能，就显得多余了

[mattdm]

@kele527's comment seems correct -- https://github.com/anthropics/skills/blob/main/skills/skill-creator/SKILL.md#frontmatter

[eovidiu]

@kele527's comment seems correct -- https://github.com/anthropics/skills/blob/main/skills/skill-creator/SKILL.md#frontmatter

Done.