Back to Skills

Plugin Auditor

jeremylongshore
Updated Today
25 views
927
113
927
View on GitHub
Developmentaiautomation

About

Plugin Auditor automatically reviews Claude Code plugins for security vulnerabilities, best practices, and compliance with repository standards. It triggers when users mention audit, security review, or best practices checks, scanning for issues like hardcoded secrets and dangerous commands. This skill uses Read, Grep, and Bash tools to perform its automated analysis.

Documentation

Plugin Auditor

Purpose

Automatically audits Claude Code plugins for security vulnerabilities, best practice violations, CLAUDE.md compliance, and quality standards - optimized for claude-code-plugins repository requirements.

Trigger Keywords

  • "audit plugin"
  • "security review" or "security audit"
  • "best practices check"
  • "plugin quality"
  • "compliance check"
  • "plugin security"

Audit Categories

1. Security Audit

Critical Checks:

  • ❌ No hardcoded secrets (passwords, API keys, tokens)
  • ❌ No AWS keys (AKIA...)
  • ❌ No private keys (BEGIN PRIVATE KEY)
  • ❌ No dangerous commands (rm -rf /, eval(), exec())
  • ❌ No command injection vectors
  • ❌ No suspicious URLs (IP addresses, non-HTTPS)
  • ❌ No obfuscated code (base64 decode, hex encoding)

Security Patterns:

# Check for hardcoded secrets
grep -r "password\s*=\s*['\"]" --exclude-dir=node_modules
grep -r "api_key\s*=\s*['\"]" --exclude-dir=node_modules
grep -r "secret\s*=\s*['\"]" --exclude-dir=node_modules

# Check for AWS keys
grep -r "AKIA[0-9A-Z]{16}" --exclude=README.md

# Check for private keys
grep -r "BEGIN.*PRIVATE KEY" --exclude=README.md

# Check for dangerous patterns
grep -r "rm -rf /" | grep -v "/var/" | grep -v "/tmp/"
grep -r "eval\s*\(" --exclude=README.md

2. Best Practices Audit

Plugin Structure:

  • ✅ Proper directory hierarchy
  • ✅ Required files present
  • ✅ Semantic versioning (x.y.z)
  • ✅ Clear, concise descriptions
  • ✅ Proper LICENSE file (MIT/Apache-2.0)
  • ✅ Comprehensive README
  • ✅ At least 5 keywords

Code Quality:

  • ✅ No TODO/FIXME without issue links
  • ✅ No console.log() in production code
  • ✅ No hardcoded paths (/home/, /Users/)
  • ✅ Uses ${CLAUDE_PLUGIN_ROOT} in hooks
  • ✅ Scripts have proper shebangs
  • ✅ All scripts are executable

Documentation:

  • ✅ README has installation section
  • ✅ README has usage examples
  • ✅ README has clear description
  • ✅ Commands have proper frontmatter
  • ✅ Agents have model specified
  • ✅ Skills have trigger keywords

3. CLAUDE.md Compliance

Repository Standards:

  • ✅ Follows plugin structure from CLAUDE.md
  • ✅ Uses correct marketplace slug
  • ✅ Proper category assignment
  • ✅ Valid plugin.json schema
  • ✅ Marketplace catalog entry exists
  • ✅ Version consistency

Skills Compliance (if applicable):

  • ✅ SKILL.md has proper frontmatter
  • ✅ Description includes trigger keywords
  • ✅ allowed-tools specified (if restricted)
  • ✅ Clear purpose and instructions
  • ✅ Examples provided

4. Marketplace Compliance

Catalog Requirements:

  • ✅ Plugin listed in marketplace.extended.json
  • ✅ Source path matches actual location
  • ✅ Version matches plugin.json
  • ✅ Category is valid
  • ✅ No duplicate plugin names
  • ✅ Author information complete

5. Git Hygiene

Repository Practices:

  • ✅ No large binary files
  • ✅ No node_modules/ committed
  • ✅ No .env files
  • ✅ Proper .gitignore
  • ✅ No merge conflicts
  • ✅ Clean commit history

6. MCP Plugin Audit (if applicable)

MCP-Specific Checks:

  • ✅ Valid package.json with @modelcontextprotocol/sdk
  • ✅ TypeScript configured correctly
  • ✅ dist/ in .gitignore
  • ✅ Proper mcp/*.json configuration
  • ✅ Build scripts present
  • ✅ No dependency vulnerabilities

7. Performance Audit

Efficiency Checks:

  • ✅ No unnecessary file reads
  • ✅ Efficient glob patterns
  • ✅ No recursive loops
  • ✅ Reasonable timeout values
  • ✅ No memory leaks (event listeners)

8. Accessibility & UX

User Experience:

  • ✅ Clear error messages
  • ✅ Helpful command descriptions
  • ✅ Proper usage examples
  • ✅ Good README formatting
  • ✅ Working demo commands

Audit Process

When activated, I will:

  1. Security Scan

    # Run security checks
    grep -r "password\|secret\|api_key" plugins/plugin-name/
    grep -r "AKIA[0-9A-Z]{16}" plugins/plugin-name/
    grep -r "BEGIN.*PRIVATE KEY" plugins/plugin-name/
    grep -r "rm -rf /" plugins/plugin-name/
    grep -r "eval\(" plugins/plugin-name/
    
  2. Structure Validation

    # Check required files
    test -f .claude-plugin/plugin.json
    test -f README.md
    test -f LICENSE
    
    # Check component directories
    ls -d commands/ agents/ skills/ hooks/ mcp/ 2>/dev/null
    
  3. Best Practices Check

    # Check for TODO/FIXME
    grep -r "TODO\|FIXME" --exclude=README.md
    
    # Check for console.log
    grep -r "console\.log" --exclude=README.md
    
    # Check script permissions
    find . -name "*.sh" ! -perm -u+x
    
  4. Compliance Verification

    # Check marketplace entry
    jq '.plugins[] | select(.name == "plugin-name")' .claude-plugin/marketplace.extended.json
    
    # Verify version consistency
    plugin_version=$(jq -r '.version' .claude-plugin/plugin.json)
    market_version=$(jq -r '.plugins[] | select(.name == "plugin-name") | .version' .claude-plugin/marketplace.extended.json)
    
  5. Generate Audit Report

Audit Report Format

🔍 PLUGIN AUDIT REPORT
Plugin: plugin-name
Version: 1.0.0
Category: security
Audit Date: 2025-10-16

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
🔒 SECURITY AUDIT
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

✅ PASSED (7/7)
- No hardcoded secrets
- No AWS keys
- No private keys
- No dangerous commands
- No command injection vectors
- HTTPS URLs only
- No obfuscated code

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
📋 BEST PRACTICES
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

✅ PASSED (10/12)
- Proper directory structure
- Required files present
- Semantic versioning
- Clear descriptions
- Comprehensive README

⚠️  WARNINGS (2)
- 3 scripts missing execute permission
  Fix: chmod +x scripts/*.sh

- 2 TODO items without issue links
  Location: commands/scan.md:45, agents/analyzer.md:67
  Recommendation: Create GitHub issues or remove TODOs

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
✅ CLAUDE.MD COMPLIANCE
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

✅ PASSED (6/6)
- Follows plugin structure
- Uses correct marketplace slug
- Proper category assignment
- Valid plugin.json schema
- Marketplace entry exists
- Version consistency

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
📊 QUALITY SCORE
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Security:        10/10 ✅
Best Practices:   8/10 ⚠️
Compliance:      10/10 ✅
Documentation:   10/10 ✅

OVERALL SCORE: 9.5/10 (EXCELLENT)

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
🎯 RECOMMENDATIONS
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Priority: MEDIUM
1. Fix script permissions (2 min)
2. Resolve TODO items (10 min)

Optional Improvements:
- Add more usage examples in README
- Include troubleshooting section
- Add GIF/video demo

✅ AUDIT COMPLETE
Plugin is production-ready with minor improvements needed.

Severity Levels

Critical (🔴):

  • Security vulnerabilities
  • Hardcoded secrets
  • Dangerous commands
  • Missing required files

High (🟠):

  • Best practice violations
  • Missing documentation
  • Broken functionality
  • Schema violations

Medium (🟡):

  • Code quality issues
  • Missing optional features
  • Performance concerns
  • UX improvements

Low (🟢):

  • Style inconsistencies
  • Minor documentation gaps
  • Nice-to-have features

Auto-Fix Capabilities

I can automatically fix:

  • ✅ Script permissions
  • ✅ JSON formatting
  • ✅ Markdown formatting
  • ✅ Version sync issues

Repository-Specific Checks

For claude-code-plugins repo:

  • Validates against CLAUDE.md standards
  • Checks marketplace integration
  • Verifies category structure
  • Ensures quality for featured plugins
  • Checks contributor guidelines compliance

Examples

User says: "Audit the security-scanner plugin"

I automatically:

  1. Run full security scan
  2. Check best practices
  3. Verify CLAUDE.md compliance
  4. Generate comprehensive report
  5. Provide recommendations

User says: "Is this plugin safe to publish?"

I automatically:

  1. Security audit (critical)
  2. Marketplace compliance
  3. Quality score calculation
  4. Publish readiness assessment

User says: "Quality review before featured status"

I automatically:

  1. Full audit (all categories)
  2. Higher quality thresholds
  3. Featured plugin requirements
  4. Recommendation: approve/reject

Quick Install

/plugin add https://github.com/jeremylongshore/claude-code-plugins-plus-skills/tree/main/plugin-auditor

Copy and paste this command in Claude Code to install this skill

GitHub 仓库

jeremylongshore/claude-code-plugins-plus-skills
Path: backups/plugin-enhancements/plugin-backups/skills-powerkit_20251019_164014/skills/plugin-auditor
aiautomationclaude-codedevopsmarketplacemcp

Related Skills

sglang

Meta

SGLang is a high-performance LLM serving framework that specializes in fast, structured generation for JSON, regex, and agentic workflows using its RadixAttention prefix caching. It delivers significantly faster inference, especially for tasks with repeated prefixes, making it ideal for complex, structured outputs and multi-turn conversations. Choose SGLang over alternatives like vLLM when you need constrained decoding or are building applications with extensive prefix sharing.

View skill

evaluating-llms-harness

Testing

This Claude Skill runs the lm-evaluation-harness to benchmark LLMs across 60+ standardized academic tasks like MMLU and GSM8K. It's designed for developers to compare model quality, track training progress, or report academic results. The tool supports various backends including HuggingFace and vLLM models.

View skill

llamaguard

Other

LlamaGuard is Meta's 7-8B parameter model for moderating LLM inputs and outputs across six safety categories like violence and hate speech. It offers 94-95% accuracy and can be deployed using vLLM, Hugging Face, or Amazon SageMaker. Use this skill to easily integrate content filtering and safety guardrails into your AI applications.

View skill

langchain

Meta

LangChain is a framework for building LLM applications using agents, chains, and RAG pipelines. It supports multiple LLM providers, offers 500+ integrations, and includes features like tool calling and memory management. Use it for rapid prototyping and deploying production systems like chatbots, autonomous agents, and question-answering services.

View skill