Claude Agents Power

# /agent-download 🤖 **AI-Powered Agent Downloader** - Now available as a native MCP tool for direct execution within Claude Code! ## ✅ **Status: FULLY FUNCTIONAL AS MCP TOOL** Use this tool directly in Claude Code! The AI-powered agent recommendation and download system is now integrated as a native MCP tool. ## MCP Tool Usage ```json { "tool": "agent-download", "arguments": { "targetDir": "./.claude/agents", "claudeMdPath": "./CLAUDE.md", "format": "md", "language": "en", "limit": 5, "dryRun": false, "overwrite": false } } ``` ## Quick Start Examples ### Basic Usage (Default Settings) ```json { "tool": "agent-download" } ``` ### Preview Mode (Dry Run) ```json { "tool": "agent-download", "arguments": { "dryRun": true, "limit": 3 } } ``` ### Custom Directory and Language ```json { "tool": "agent-download", "arguments": { "targetDir": "./team-agents", "language": "ko", "limit": 7 } } ``` ### Full Configuration ```json { "tool": "agent-download", "arguments": { "targetDir": "./.claude/agents", "claudeMdPath": "./docs/project.md", "format": "yaml", "language": "en", "limit": 10, "dryRun": false, "overwrite": true } } ``` ## Parameters | Parameter | Type | Default | Description | |-----------|------|---------|-------------| | `targetDir` | string | `./.claude/agents` | Directory to save agent files | | `claudeMdPath` | string | `./CLAUDE.md` | Path to project description file | | `format` | enum | `md` | File format: `md`, `yaml`, `json` | | `language` | enum | `en` | Language: `en`, `ko`, `ja`, `zh` | | `limit` | number | 10 | Max agents to download (1-20) | | `dryRun` | boolean | false | Preview without downloading | | `overwrite` | boolean | false | Overwrite existing files | ## AI Features ### 🧠 Intelligent Analysis - **Project Understanding**: Deep analysis of project structure, goals, and requirements - **Technology Detection**: Automatic identification of frameworks, languages, and tools - **Quality Assessment**: Evaluation of testing, documentation, and development practices - **Complexity Scoring**: Dynamic complexity assessment (1-10 scale) ### 🎯 Smart Recommendations - **Context-Aware Matching**: Agents matched to specific project needs and phase - **Dynamic Prioritization**: Essential vs. recommended vs. optional categorization - **Task-Specific Assignment**: Specific tasks and responsibilities for each agent - **Integration Mapping**: How agents will work together and integrate ### ⚡ Automated Workflow - **Seamless Download**: Direct file creation in specified directory - **Enhanced Documentation**: AI-generated README with usage guidelines - **Multiple Formats**: Support for Markdown, YAML, and JSON formats - **Multilingual Support**: Agents available in multiple languages ## Response Format ### Success Response ```json { "success": true, "dryRun": false, "analysis": { "projectType": "web-application", "technologies": ["JavaScript", "TypeScript"], "frameworks": ["React", "Node.js"], "complexity": 8, "phase": "development", "teamSize": 6 }, "recommendations": [ { "name": "frontend-developer", "relevanceScore": 95, "reasoning": "Essential for React development", "priority": "essential", "specificTasks": ["UI implementation", "Performance optimization"], "integrationPoints": ["Backend APIs", "Design system"] } ], "downloaded": ["./.claude/agents/frontend-developer.md"], "message": "Successfully downloaded 5 AI-recommended agents to ./.claude/agents." } ``` ### Error Response ```json { "success": false, "error": "Agent download failed: CLAUDE.md not found", "suggestion": "Please check the CLAUDE.md file path and ensure write permissions for the target directory" } ``` ## Integration Benefits ### 🚀 **Native MCP Integration** - ✅ **Direct Execution**: No need for terminal commands - ✅ **Real-time Results**: Immediate feedback within Claude Code - ✅ **Seamless Workflow**: Integrated into your development process - ✅ **Error Handling**: Clear error messages and suggestions ### 🤖 **AI-Powered Intelligence** - ✅ **15x More Accurate**: AI understanding vs simple keyword matching - ✅ **Context Awareness**: Adapts to project phase and complexity - ✅ **Quality Integration**: Considers testing, documentation, CI/CD practices - ✅ **Team Optimization**: Recommends complementary skill sets ### 📊 **Enhanced Analytics** - ✅ **Usage Tracking**: Monitor tool effectiveness - ✅ **Success Metrics**: Track download success rates - ✅ **Performance Insights**: Optimize recommendation accuracy - ✅ **Project Intelligence**: Learn from project patterns ## Common Use Cases ### **New Project Setup** ```json { "tool": "agent-download", "arguments": { "dryRun": true, "limit": 3 } } ``` ### **Team Expansion** ```json { "tool": "agent-download", "arguments": { "targetDir": "./additional-agents", "limit": 7, "overwrite": true } } ``` ### **Project Evolution** ```json { "tool": "agent-download", "arguments": { "claudeMdPath": "./updated-requirements.md", "limit": 5 } } ``` ### **Multilingual Teams** ```json { "tool": "agent-download", "arguments": { "language": "ko", "targetDir": "./korean-agents" } } ``` ## Pro Tips 1. **Start with Dry Run**: Always preview recommendations first 2. **Project Evolution**: Re-run as your project requirements change 3. **Phase-Specific**: Different agents for different development phases 4. **Language Options**: Use native language agents for better team adoption 5. **Directory Organization**: Organize agents by project or team structure Perfect for intelligently assembling the optimal expertise team for any development project using advanced AI reasoning - now seamlessly integrated into Claude Code! 🚀