{"id":28814767,"url":"https://github.com/krzemienski/enhanced-claude-code","last_synced_at":"2026-07-01T19:31:45.615Z","repository":{"id":299123005,"uuid":"1002123175","full_name":"krzemienski/enhanced-claude-code","owner":"krzemienski","description":"Advanced Claude Code Tutorial","archived":false,"fork":false,"pushed_at":"2025-07-03T22:03:52.000Z","size":899,"stargazers_count":4,"open_issues_count":1,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2025-07-03T22:28:30.042Z","etag":null,"topics":["ai-code-generation","anthropic","claude","claude-code","claude-code-sdk","opus4","vibe-coding"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/krzemienski.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null}},"created_at":"2025-06-14T19:05:00.000Z","updated_at":"2025-06-26T11:17:40.000Z","dependencies_parsed_at":"2025-06-14T20:44:23.477Z","dependency_job_id":null,"html_url":"https://github.com/krzemienski/enhanced-claude-code","commit_stats":null,"previous_names":["krzemienski/enhanced-claude-code"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/krzemienski/enhanced-claude-code","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/krzemienski%2Fenhanced-claude-code","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/krzemienski%2Fenhanced-claude-code/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/krzemienski%2Fenhanced-claude-code/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/krzemienski%2Fenhanced-claude-code/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/krzemienski","download_url":"https://codeload.github.com/krzemienski/enhanced-claude-code/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/krzemienski%2Fenhanced-claude-code/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":35020870,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-26T15:22:16.424Z","status":"online","status_checked_at":"2026-07-01T02:00:05.325Z","response_time":130,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["ai-code-generation","anthropic","claude","claude-code","claude-code-sdk","opus4","vibe-coding"],"created_at":"2025-06-18T16:02:01.547Z","updated_at":"2026-07-01T19:31:45.610Z","avatar_url":"https://github.com/krzemienski.png","language":"Python","funding_links":[],"categories":["HarmonyOS"],"sub_categories":["Windows Manager"],"readme":"# Claude Code Builder v3.0: AI-Driven Autonomous Project Builder\n\n\u003e **Transform Claude Code into an AI-powered architect that dynamically plans and builds complex projects with intelligent phase generation and comprehensive testing**\n\n## 🎯 What This Actually Does\n\nThis repository demonstrates the most advanced techniques for orchestrating Claude Code with:\n- **AI-driven phase planning** - Claude analyzes specs and creates optimal build strategies\n- **Real-time streaming output** - Visual tool tracking and cost monitoring\n- **Comprehensive functional testing** - NO DRY RUNS, only real builds (10-30 minutes)\n- **Enhanced context management** - Persistent memory with error logging\n- **Dynamic MCP server discovery** - Automatic detection and recommendations\n- **Intelligent recovery** - Checkpoint every 5 minutes with full context\n\n**The result**: A v3.0 builder that uses AI to plan phases, executes with rich visual feedback, and validates through real functional testing in 25-45 minutes.\n\n## 🚀 What's New in v3.0\n\n### AI-Driven Planning\n- Claude analyzes your project specification to create optimal phases\n- Dynamically generates 15-20 phases based on complexity\n- Intelligent dependency resolution and task breakdown\n- Risk assessment and mitigation strategies\n\n### Enhanced Visual Feedback\n- Real-time tool usage tracking with visual indicators\n- Streaming JSON parsing for immediate feedback\n- Cost tracking displays (session and total)\n- Progress bars with phase completion tracking\n\n### Comprehensive Testing (NO DRY RUNS)\n- All tests execute real builds that create actual files\n- Multi-stage validation process\n- Automated test project creation and execution\n- Performance benchmarking and cost analysis\n- **Critical**: Tests WILL take 10-30 minutes - this is NORMAL\n\n## 🔥 The Problem We're Solving\n\nWhen you ask Claude Code to build a complex project in one shot:\n```bash\nclaude \"Build me a Python package with 7 AI agents, MCP integration, and Rich UI\"\n```\n\n**What happens**:\n- ❌ Claude loses context after ~20 files\n- ❌ Forgets architectural decisions made earlier\n- ❌ Creates incompatible interfaces between components\n- ❌ No way to resume if interrupted\n- ❌ Success rate: ~60%\n\n## 💡 Our Solution: Intelligent Orchestration\n\n### How v3.0 Works\n\n\u003e **Note**: If the Mermaid diagrams below are not rendering on GitHub, you can:\n\u003e - View them locally by cloning the repo\n\u003e - Use a Mermaid live editor: https://mermaid.live\n\u003e - Install a browser extension for Mermaid rendering\n\n```mermaid\nflowchart TD\n    A[Project Specification] --\u003e B[v3.0 Builder Script]\n    B --\u003e C{AI Planning Phase}\n    \n    C --\u003e D[Claude Analyzes Spec]\n    D --\u003e E[Generate Optimal Phases]\n    E --\u003e F[Create Task Breakdown]\n    F --\u003e G[build-phases-v3.json]\n    \n    G --\u003e H{Dynamic Execution}\n    H --\u003e I[Phase 1-N: Build]\n    \n    I --\u003e J[Real-time Streaming]\n    J --\u003e K[Visual Tool Tracking]\n    K --\u003e L[Cost Monitoring]\n    \n    L --\u003e M[Phase Complete]\n    M --\u003e N{Next Phase?}\n    N --\u003e|Yes| I\n    N --\u003e|No| O[Functional Testing]\n    \n    O --\u003e P[NO DRY RUNS]\n    P --\u003e Q[Real Build Test]\n    Q --\u003e R[10-30 Minutes]\n    R --\u003e S[Validation]\n    \n    S --\u003e T[Build Complete]\n    \n    style A fill:#f9f,stroke:#333,stroke-width:2px\n    style C fill:#ff9,stroke:#333,stroke-width:2px\n    style O fill:#9ff,stroke:#333,stroke-width:2px\n    style T fill:#9f9,stroke:#333,stroke-width:2px\n```\n\n## 📊 Real Execution: What Actually Happens\n\n### Phase 0: AI-Driven Planning\n```\n═══ PHASE ═══ Phase 0: AI-Driven Planning\n[INFO] Invoking Claude to analyze specification and create build plan...\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n\n🔧 [TOOL USE] filesystem__write_file\n   ID: msg_01RfKYiQ1CjtJ8vBaYBRX6Xe\n✓ [TOOL RESULT] Completed\n\n💬 [CLAUDE] I've analyzed the specification and created an optimal 18-phase build plan\n🔧 [TOOL USE] filesystem__write_file\n   ID: msg_013xBgsp7jSifTs2YjpjeSeu\n✓ [TOOL RESULT] Completed\n\n[SUCCESS] ✅ AI planning completed successfully\n[INFO] AI planned 18 phases for optimal build\n```\n\n### Dynamic Phase Execution with Visual Feedback\n```\n📊 Progress: 5/18 phases complete\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n\n═══ PHASE ═══ Phase 6: Enhanced Research System\n[INFO] Objective: Implement 7 specialized research agents\n[INFO] Starting execution with Claude Code SDK\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n\n🔧 [TOOL USE] mem0__search-memory\n   ID: msg_01ESG9GKgJYwsU6TZbWBC4wj\n💬 [CLAUDE] Retrieving base agent implementation from previous phases\n\n🔧 [TOOL USE] filesystem__write_file\n   ID: msg_0122p7cJ7PrpwAmsABzhSffK\n✓ [TOOL RESULT] Completed\n\n[COST] Phase 6 - Session: $0.42, Total: $2.85\n[INFO] Phase 6 completed in 127 seconds\n```\n\n### Comprehensive Functional Testing (NO DRY RUNS)\n```\n═══ PHASE ═══ Final Phase: Comprehensive Functional Testing\n[WARNING] ⚠️  This phase will take 10-30 minutes. This is NORMAL.\n\n🧪 FUNCTIONAL TESTING STAGES:\n  1️⃣  Installation verification\n  2️⃣  CLI functionality testing\n  3️⃣  Full project build (10-30 minutes)\n  4️⃣  Results validation\n\n[INFO] Starting functional testing with extended timeout (30 minutes)\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n\n💬 [CLAUDE] Starting Stage 1: Installation Test\n🔧 [TOOL USE] Bash\n   ID: msg_01CT9JkuLcfPzERYSYL7md6p\n\n=== Build Progress Check ===\nFiles created: 0\nPhases completed: 0\nErrors: 0\nRecent activity:\nWaiting for output...\n=========================\n\n[... continues monitoring for 10-30 minutes ...]\n\n=== Build Progress Check ===\nFiles created: 47\nPhases completed: 12\nErrors: 0\nRecent activity:\n[PHASE COMPLETE] ✅ Phase 12: Documentation\nCreated: claude_code_builder/README.md\nTool: Write Creating final configuration\nSession cost: $0.38\n=========================\n```\n\n## 🚀 Key Features Demonstrated\n\n### 1. **AI-Driven Phase Planning**\n```json\n{\n  \"version\": \"3.0.0\",\n  \"total_phases\": 18,\n  \"phases\": [\n    {\n      \"number\": 1,\n      \"name\": \"Foundation and Architecture\",\n      \"objective\": \"Set up project structure with v3.0 enhancements\",\n      \"tasks\": [\n        \"Create modular package structure\",\n        \"Initialize development environment\",\n        \"Set up comprehensive error handling\",\n        \"Configure logging and telemetry\",\n        \"Create base exception classes\"\n      ],\n      \"complexity\": 3,\n      \"dependencies\": [],\n      \"estimated_time\": \"3-5 minutes\"\n    }\n  ]\n}\n```\n\n### 2. **Real-time Streaming Output**\nThe v3.0 builder parses Claude Code's streaming JSON output to provide:\n- Visual tool usage indicators\n- Real-time cost tracking\n- Progress updates\n- Error detection\n- File creation monitoring\n\n### 3. **NO DRY RUNS - Real Testing Only**\n```bash\n# v3.0 NEVER uses --dry-run\n# All tests create real files and validate actual functionality\n\nclaude-code-builder test-api-spec.md \\\n    --output-dir ./test-output \\\n    --enable-research \\\n    --verbose \\\n    --stream-output  # Real execution, not a dry run!\n```\n\n## 🛠️ Installation \u0026 Usage\n\n### Prerequisites\n```bash\n# Required tools\nnpm install -g @anthropic-ai/claude-code  # Claude CLI\nbrew install jq                           # JSON processor (macOS)\n# or\napt-get install jq                        # Ubuntu/Debian\n\n# Python 3.8+ required\npython --version\n```\n\n### Basic Usage\n```bash\n# Clone the repository\ngit clone \u003crepository-url\u003e\ncd enhanced-claude-code\n\n# Run the v3 builder\n./builder-claude-code-builder-v3.sh\n\n# The process will:\n# 1. Use AI to plan optimal phases (2-5 minutes)\n# 2. Execute each phase (20-35 minutes)\n# 3. Run comprehensive testing (10-30 minutes)\n# Total time: 25-45 minutes\n```\n\n### Understanding the Process\n\n#### Phase 0: AI Planning\n```\n[PHASE] Phase 0: AI-Driven Planning\n- Claude analyzes the specification\n- Creates optimal phase breakdown\n- Generates detailed task lists\n- Produces build-phases-v3.json\n```\n\n#### Dynamic Phases (1-N)\n```\n[PHASE] Phase 1: Foundation and Architecture\n[PHASE] Phase 2: Enhanced Data Models\n... (AI determines optimal phases)\n[PHASE] Phase N: Documentation and Polish\n```\n\n#### Final Phase: Functional Testing\n```\n[PHASE] Final Phase: Comprehensive Functional Testing\n[WARNING] This phase will take 10-30 minutes. This is NORMAL.\n\nStage 1: Installation verification\nStage 2: CLI testing  \nStage 3: Full functional test (LONG RUNNING - NO DRY RUN)\nStage 4: Validation\n```\n\n## 💰 Cost Expectations\n\nTypical v3.0 build costs:\n- AI Planning Phase: $0.20 - $0.50\n- Build Phases: $3.00 - $5.00\n- Functional Testing: $1.00 - $2.00\n- **Total: $4.20 - $7.50**\n\nCost breakdown by category:\n- Claude Code execution: 40-50%\n- Research APIs: 20-30%\n- Analysis \u0026 Planning: 30-40%\n\n## 📈 Performance Comparison\n\n| Feature | v2.3.0 | v3.0 |\n|---------|---------|------| \n| Phase Planning | Static 15 phases | AI-generated optimal phases |\n| Testing | Basic validation | Comprehensive functional testing (NO DRY RUNS) |\n| Monitoring | Basic progress | Real-time streaming with analytics |\n| Visual Output | Standard logging | Rich visual feedback with tool tracking |\n| Build Time | ~20-30 minutes | 25-45 minutes (includes testing) |\n| Success Rate | ~85% | ~95% |\n\n## 🔧 Advanced Configuration\n\n### Custom MCP Servers\nThe v3.0 builder automatically discovers and configures MCP servers:\n```json\n{\n    \"mcpServers\": {\n        \"filesystem\": {\n            \"command\": \"npx\",\n            \"args\": [\"-y\", \"@modelcontextprotocol/server-filesystem\", \".\"]\n        },\n        \"memory\": {\n            \"command\": \"npx\",\n            \"args\": [\"-y\", \"@modelcontextprotocol/server-memory\"]\n        },\n        \"sequential-thinking\": {\n            \"command\": \"npx\",\n            \"args\": [\"-y\", \"@modelcontextprotocol/server-sequential-thinking\"]\n        }\n    }\n}\n```\n\n### Environment Variables\n```bash\nexport ANTHROPIC_API_KEY=\"your-key\"  # Required for research features\nexport CLAUDE_MODEL=\"claude-3-5-sonnet-20241022\"  # Override default model\n```\n\n## 🐛 Troubleshooting\n\n### Build Seems Stuck\n**This is usually normal!** Check if:\n- Files are still being created: `find . -name \"*.py\" -mmin -5`\n- Log is still growing: `ls -la *.log`\n- Process is still running: `ps aux | grep claude`\n\n### Memory Issues\nIf you see memory-related errors:\n```bash\n# Clear MCP memory state\nrm -rf .memory_keys/\nrm .memory_keys.json\n```\n\n### Recovery from Failure\nThe build automatically checkpoints every 5 minutes:\n```bash\n# Resume from checkpoint\n./builder-claude-code-builder-v3.sh --resume\n```\n\n## 📚 Repository Structure\n\n```\nenhanced-claude-code/\n├── builder-claude-code-builder-v3.sh    # Main v3.0 orchestrator\n├── prompt.md                            # v2.3.0 project specification\n├── phases.md                            # Static phase definitions (v2.x)\n├── tasks.md                             # Detailed task breakdowns (v2.x)\n├── instructions.md                      # Implementation guidelines\n├── test-manager-v3.sh                   # Comprehensive test framework\n├── testing-instructions-v3.md           # NO DRY RUN testing guide\n└── README.md                            # This file\n```\n\n## 🤝 Contributing\n\nWhen contributing to v3.0:\n1. Ensure AI planning phase remains flexible\n2. Maintain comprehensive testing requirements (NO DRY RUNS)\n3. Preserve real-time monitoring capabilities\n4. Document any changes to timing expectations\n\n## 📖 Documentation\n\n- [Full v3.0 Specification](claude-code-builder-v3-spec.md)\n- [Testing Instructions](testing-instructions-v3.md)\n- [AI Planning Prompt](ai-planning-prompt.md)\n- [Legacy v2.3.0 Documentation](prompt.md)\n\n## 💡 Key Takeaways\n\n1. **AI Planning Works**: Letting Claude analyze and plan phases produces better results than static definitions\n2. **Visual Feedback Matters**: Real-time tool tracking helps debug and understand execution\n3. **Real Testing Only**: NO DRY RUNS - actual builds catch real issues\n4. **Time Investment Pays Off**: 25-45 minute builds with testing produce 95% success rates\n5. **Context is King**: Enhanced memory and recovery make complex builds possible\n\n---\n\n**Remember**: The functional test WILL take 10-30 minutes. This is NORMAL and EXPECTED behavior for v3.0!\n\n🤖 Built with Claude Code v3.0","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fkrzemienski%2Fenhanced-claude-code","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fkrzemienski%2Fenhanced-claude-code","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fkrzemienski%2Fenhanced-claude-code/lists"}