daffy0208__ai-dev-standards/scripts/test-plugins.cjs

#!/usr/bin/env node
/**
 * Plugin Verification Test Script
 *
 * Tests all installed Claude Code plugins from the marketplace:
 * - Verifies plugin installation
 * - Tests MCP server plugins are running
 * - Validates agent skills are accessible
 * - Checks plugin relationships with existing skills
 * - Reports integration status
 */

const fs = require('fs');
const path = require('path');
const { execSync } = require('child_process');

const ROOT = path.resolve(__dirname, '..');
const COLORS = {
  reset: '\x1b[0m',
  red: '\x1b[31m',
  green: '\x1b[32m',
  yellow: '\x1b[33m',
  cyan: '\x1b[36m',
  bright: '\x1b[1m',
};

function colorize(text, color) {
  return `${COLORS[color]}${text}${COLORS.reset}`;
}

function readJSON(filePath) {
  try {
    return JSON.parse(fs.readFileSync(filePath, 'utf8'));
  } catch (error) {
    return null;
  }
}

console.log(colorize('\n=== CLAUDE CODE PLUGIN VERIFICATION ===\n', 'bright'));

// Expected plugins from installation
const expectedPlugins = [
  {
    name: 'ai-ml-engineering-pack',
    type: 'ai-instruction',
    description: 'RAG, LLM integration, vector DB patterns',
    skillRelations: ['rag-implementer', 'knowledge-base-manager', 'knowledge-graph-builder']
  },
  {
    name: 'skills-powerkit',
    type: 'agent-skill',
    description: 'Plugin management automation',
    skillRelations: ['all-skills']
  },
  {
    name: 'devops-automation-pack',
    type: 'ai-instruction',
    description: 'CI/CD, Docker, Kubernetes patterns',
    skillRelations: ['deployment-advisor']
  },
  {
    name: 'workflow-orchestrator',
    type: 'mcp-server',
    description: 'DAG-based workflow automation',
    skillRelations: ['orchestration-planner', 'multi-agent-architect']
  },
  {
    name: 'project-health-auditor',
    type: 'mcp-server',
    description: 'Code health analysis',
    skillRelations: ['quality-auditor', 'system-diagnostician']
  },
  {
    name: 'domain-memory-agent',
    type: 'mcp-server',
    description: 'Semantic search knowledge base',
    skillRelations: ['rag-implementer', 'knowledge-base-manager']
  },
  {
    name: 'security-pro-pack',
    type: 'ai-instruction',
    description: 'OWASP auditing, threat modeling',
    skillRelations: ['security-engineer']
  },
  {
    name: 'agent-context-manager',
    type: 'agent-skill',
    description: 'Auto-detect AGENTS.md files',
    skillRelations: ['context-preserver']
  },
  {
    name: 'fullstack-starter-pack',
    type: 'ai-instruction',
    description: 'React, API scaffolding patterns',
    skillRelations: ['frontend-builder', 'api-designer']
  },
  {
    name: 'design-to-code',
    type: 'mcp-server',
    description: 'Design to React/Vue/Svelte',
    skillRelations: ['frontend-builder', 'design-system-architect']
  }
];

const results = {
  installed: [],
  missing: [],
  mcpServers: [],
  aiInstructions: [],
  agentSkills: [],
  errors: []
};

console.log(colorize('1. Checking Plugin Installation Status...', 'cyan'));

// Check if plugins are installed by looking for Claude Code plugin directory
// Note: Actual plugin location depends on Claude Code's installation directory
// This is a placeholder - actual verification would require Claude Code API

expectedPlugins.forEach(plugin => {
  console.log(`   Checking: ${plugin.name}`);

  // Since we can't directly verify Claude Code plugins without API access,
  // we'll document what should be verified
  switch(plugin.type) {
    case 'mcp-server':
      results.mcpServers.push(plugin);
      console.log(colorize(`   → ${plugin.name}: MCP Server (should be running as Node.js process)`, 'yellow'));
      break;
    case 'ai-instruction':
      results.aiInstructions.push(plugin);
      console.log(colorize(`   → ${plugin.name}: AI Instruction Template (loaded on demand)`, 'green'));
      break;
    case 'agent-skill':
      results.agentSkills.push(plugin);
      console.log(colorize(`   → ${plugin.name}: Agent Skill (auto-activates)`, 'green'));
      break;
  }
});

console.log(colorize('\n2. Plugin Type Distribution...', 'cyan'));
console.log(`   MCP Servers: ${results.mcpServers.length}`);
console.log(`   AI Instructions: ${results.aiInstructions.length}`);
console.log(`   Agent Skills: ${results.agentSkills.length}`);
console.log(`   Total: ${expectedPlugins.length}`);

console.log(colorize('\n3. Checking Skill Relationships...', 'cyan'));
const skillRegistry = readJSON(path.join(ROOT, 'meta/skill-registry.json'));
if (skillRegistry) {
  const registeredSkills = skillRegistry.skills.map(s => s.name);

  expectedPlugins.forEach(plugin => {
    console.log(`\n   ${plugin.name}:`);
    console.log(`   → Enhances: ${plugin.skillRelations.join(', ')}`);

    plugin.skillRelations.forEach(skillName => {
      if (skillName === 'all-skills') {
        console.log(colorize(`     ✓ Applies to all ${registeredSkills.length} skills`, 'green'));
      } else if (registeredSkills.includes(skillName)) {
        console.log(colorize(`     ✓ ${skillName} exists`, 'green'));
      } else {
        console.log(colorize(`     ✗ ${skillName} not found in registry`, 'red'));
        results.errors.push(`Missing skill: ${skillName} (required by ${plugin.name})`);
      }
    });
  });
}

console.log(colorize('\n4. Integration Recommendations...', 'cyan'));
console.log(`
   Next Steps:

   1. Verify MCP Servers Are Running:
      - workflow-orchestrator
      - project-health-auditor
      - domain-memory-agent
      - design-to-code

      Check with: ps aux | grep "mcp"

   2. Test AI Instruction Templates:
      - Try prompts that should trigger the plugins
      - Example: "Help me design a RAG system" (should use ai-ml-engineering-pack)

   3. Verify Agent Skills Auto-Activation:
      - skills-powerkit should auto-activate for plugin management
      - agent-context-manager should detect AGENTS.md files

   4. Update Plugin Registry:
      - Run: node scripts/update-plugin-registry.cjs
      - This will add plugins to meta/plugin-registry.json

   5. Update Relationship Mappings:
      - Run: node scripts/update-relationships.cjs
      - This will link plugins to existing skills

   6. Re-run System Validation:
      - Run: node scripts/validate-complete-system.cjs
      - Verify no new issues introduced
`);

console.log(colorize('\n=== VERIFICATION SUMMARY ===\n', 'bright'));

if (results.errors.length === 0) {
  console.log(colorize('✓ All expected plugins documented', 'green'));
  console.log(colorize(`✓ ${results.mcpServers.length} MCP Servers to verify`, 'green'));
  console.log(colorize(`✓ ${results.aiInstructions.length} AI Instruction Templates ready`, 'green'));
  console.log(colorize(`✓ ${results.agentSkills.length} Agent Skills to test`, 'green'));
  console.log('');
  console.log(colorize('Run the integration scripts listed above to complete setup.', 'cyan'));
  process.exit(0);
} else {
  console.log(colorize(`✗ ${results.errors.length} issues found:`, 'red'));
  results.errors.forEach(err => console.log(colorize(`  - ${err}`, 'red')));
  console.log('');
  process.exit(1);
}