A community-driven registry for the Claude Code ecosystem. Not affiliated with Anthropic.
Are you the author? Sign in to claim
Multi-agent orchestration for Claude Code with 15-30% token optimization, self-improving agents, and automatic verificat
Coordinate multiple AI agents for parallel task execution, intelligent task distribution, and seamless Claude Code integration.
🚀 Quick Install • Features • Documentation • Contributing • License
Created: August 20, 2025
Lines of Code: 335,998+ in src/ folder alone
Test Coverage: 184 comprehensive tests
🔒 Secure NPM Installation (Recommended)No shell scripts, no security concerns, full transparency. hljs language-bash What this does:
✅ Interactive menu to choose installation path |
/monitor command accessible from anywhere--dangerously-skip-permissions needed# Create projects without any flags!
claude /create "a todo app with React"
# Monitor all agents in real-time
/monitor start
# Configure settings
claude /configure-agentwise
# Update to latest version (NEW!)
/update-agentwise
NPM Update (Recommended):
# Run the installer again - it will detect and update existing installations
npm create agentwise@latest
# Choose "Update existing installation" when prompted
Manual Update (if preferred):
cd ~/agentwise
git pull origin main
npm install --legacy-peer-deps
📖 Complete Installation Guide | 🔧 Manual Installation
Agentwise is a comprehensive development platform that transforms project creation through intelligent automation. Built as an extension for Claude Code, it provides complete end-to-end project setup including requirements planning, database integration, GitHub configuration, and automated security protection.
Agentwise now operates without requiring the dangerous permissions flag:
# Before: Required dangerous flag
claude --dangerously-skip-permissions /create my-app
# Now: Works without the flag!
claude /create my-app
/configure-agentwise # Interactive setup wizard
/configure-agentwise permissions # Configure permission handling
/configure-agentwise workspace # Set sandbox boundaries
Upload and process documents directly in Claude Code:
/upload requirements.pdf spec # Convert PDF to project specs
/upload design-brief.docx context # Use Word docs as project context
/upload technical-docs.pdf # Process any document format
Direct integration with Figma Dev Mode MCP Server for seamless design-to-code:
/figma connect # Connect to Figma desktop
/figma generate Button # Generate component from selection
/figma sync # Sync design tokens & components
/figma tokens ./tokens.json # Export design variables
/figma image ./screenshot.png # Capture design screenshots
Enhanced image processing and visual context understanding:
/image # Visual file browser
/upload screenshot.png # Process UI screenshots
/upload mockup.jpg # Convert mockups to code
Clone and customize websites with AI assistance:
/clone-website https://example.com exact # 1:1 replica
/clone-website https://site.com similar # Keep style, change brand
Before Agentwise:
With Agentwise:
These features extend Claude Code's capabilities for more comprehensive development workflows.
Agentwise implements a dual-context system for better project organization:
Based on OpenAI's AGENTS.md specification, this provides a universal interface for AI tools:
Our proprietary real-time context awareness system:
AGENTS.md → Universal AI Interface (any tool can read)
+
CodebaseContextManager → Deep, Real-Time Understanding (Agentwise exclusive)
=
Comprehensive Context System: Universal compatibility + deep awareness
Note: Our approach combines universal compatibility (AGENTS.md) with real-time codebase understanding, making it particularly well-suited for multi-agent workflows.
🤖 Multi-Agent Orchestration
💎 Agentwise Token Optimization System✅ REALISTIC PERFORMANCE: 15-30% token reduction through intelligent optimization Our dual optimization system provides meaningful cost savings: Context Sharing (15-20% typical reduction):
Smart Caching (10-15% additional reduction):
Combined Systems: 15-30% total reduction in real-world usage |
🛠️ Advanced Development Tools
|
🚀 New Enhanced Features
|
🎮 Claude Code Integration
|
Agentwise addresses Claude Code limitations and extends its capabilities:
| Claude Code Limitation | Agentwise Solution |
|---|---|
| Can't upload files | ✅ Full document, image, and design file support |
| Can't process PDFs | ✅ PDF → Project specs conversion |
| Can't read Figma files | ✅ Figma → Code generation |
| Limited image context | ✅ Full visual understanding |
| Single agent execution | ✅ Multiple parallel agents |
| No task organization | ✅ Structured project and task management |
| No local model support | ✅ Ollama, LM Studio integration |
Note: Optimization benefits only apply to multi-agent scenarios. Single agents see no benefit or may use slightly more tokens due to overhead.
graph TB
subgraph "❌ Without Agentwise"
A1[Single User/Agent<br/>10,000 tokens]
A1 -->|Full Context| API1[Claude API<br/>💰 $0.15]
B1[5 Agents Traditional<br/>50,000 tokens]
B1 -->|5x Full Context| API2[Claude API<br/>💰 $0.75]
C1[10 Agents Traditional<br/>100,000 tokens]
C1 -->|10x Full Context| API3[Claude API<br/>💰 $1.50]
end
subgraph "✅ With Agentwise Optimization"
D1[Single User<br/>10,000 tokens]
D1 -->|No benefit| API4[Claude API<br/>💰 $0.15]
E1[5 Agents Optimized<br/>42,500 tokens]
E1 -->|Shared Context| SC1[Context<br/>Server]
SC1 -->|15% less| API5[Claude API<br/>💰 $0.64]
F1[10 Agents Optimized<br/>77,000 tokens]
F1 -->|Shared Context| SC2[Context<br/>Server]
SC2 -->|23% less| API6[Claude API<br/>💰 $1.16]
end
style A1 fill:#ff6b6b,color:#fff
style B1 fill:#ff6b6b,color:#fff
style C1 fill:#ff6b6b,color:#fff
style D1 fill:#ffd93d,color:#000
style E1 fill:#6bcf7f,color:#fff
style F1 fill:#6bcf7f,color:#fff
style SC1 fill:#4dabf7,color:#fff
style SC2 fill:#4dabf7,color:#fff
| System | Token Reduction | Status | Empirical Results |
|---|---|---|---|
| Context Sharing | 10-20% | ✅ Measured | Reduces duplicate context |
| Smart Caching | 5-10% | ✅ Measured | Avoids redundant processing |
| Combined Systems | 15-30% | ✅ Measured | Varies by project complexity |
| Agent Accuracy | +10-15% | ✅ Observed | Improved with context |
| Bug Prevention | 20-30% | ✅ Observed | Better coordination |
| Dev Speed | +15-25% | ✅ Observed | Parallel processing |
| Scenario | Agents | Traditional | Optimized | Actual Reduction | Notes |
|---|---|---|---|---|---|
| Simple Task | 1 | 10,000 | 11,500 | -15% | Overhead exceeds benefit |
| Small Project | 5 | 50,000 | 42,500 | 15% | Modest savings |
| Full Project | 10 | 100,000 | 77,000 | 23% | Good for complex tasks |
| Enterprise | 20 | 200,000 | 150,000 | 25% | Best for large projects |
All results verified through comprehensive testing - see test files for details
sequenceDiagram
participant PM as Project Manager
participant CS as Context Server<br/>(Port 3003)
participant A1 as Frontend Agent
participant A2 as Backend Agent
participant A3 as Database Agent
participant API as Claude API
Note over PM,API: 🚀 Initial Setup (One-time)
PM->>CS: Upload full project context (1,344 bytes)
CS->>CS: Store and version context
CS-->>PM: Context ID: project-123
Note over PM,API: 👥 Agent Execution (Optimized)
A1->>CS: Get context reference
CS-->>A1: Context hash + ref (50 bytes)
A1->>API: Task + context ref (250 bytes total)
A2->>CS: Get context reference
CS-->>A2: Context hash + ref (50 bytes)
A2->>API: Task + context ref (250 bytes total)
A3->>CS: Get context reference
CS-->>A3: Context hash + ref (50 bytes)
A3->>API: Task + context ref (250 bytes total)
Note over PM,API: 🔄 Updates (Differential Only)
A1->>CS: Send changes only (50 bytes)
CS->>A2: Broadcast diff
CS->>A3: Broadcast diff
Note over PM,API: ✅ Total: 5,460 bytes vs 8,400 bytes
Note over PM,API: 💰 Saved: 2,940 tokens (35% reduction)
✅ Note: Agentwise now works without requiring the --dangerously-skip-permissions flag through automatic permission handling and sandboxed execution.
Secure installation without shell scripts:
# Interactive installer - works on all platforms
npm create agentwise@latest
# Alternative package managers
yarn create agentwise
pnpm create agentwise
What you get:
For advanced users who prefer manual setup:
# 1. Clone the repository
git clone https://github.com/VibeCodingWithPhil/agentwise.git
cd ~/agentwise
# 2. Install dependencies
npm install
# 3. Build the project
npm run build
# 4. Create workspace directory
mkdir -p workspace
# 5. Install global monitor command
npm run monitor:install
# 6. Start Agentwise
node dist/index.js
# 7. In a new terminal, test without any flags!
claude /create "your first project"
Remove the need for --dangerously-skip-permissions:
# Run the configuration wizard
claude /configure-agentwise
# Or create config file directly at ~/.agentwise-config.json
{
"permissions": {
"bypassEnabled": true,
"safetyMode": "moderate",
"autoResponse": true
},
"workspace": {
"enableSandbox": true,
"restrictPaths": true,
"allowedPaths": ["~/agentwise/workspace"]
}
}
Benefits of Manual Installation:
Agentwise provides 45 powerful commands organized by category:
/create <idea> # Create project with smart agent selection
/create-plan <idea> # Collaborative planning mode
/create-project <idea> # Complete setup wizard with all features
/projects # List and select active projects
/project-status # View current project status
/task <feature> # Add features to active project
/task-dynamic <feature> # Dynamic task distribution
/task-plan <feature> # Plan tasks before execution
/task-import <source> # Import tasks from external sources
/init-import # Import external project
/clone-website <url> # Clone and customize websites
/requirements <idea> # Generate project requirements
/requirements-enhance # Enhance existing requirements
/requirements-visualize # Create visual HTML specifications
/requirements-to-tasks # Convert requirements to tasks
/database-wizard # Interactive database setup wizard
/database-setup # Quick database configuration
/database-connect # Connect to existing database
/enable-protection # Enable automated backup and security
/protection-status # View real-time protection status
/security-review # Comprehensive security analysis
/security-report # Generate detailed security report
/rollback # Rollback to previous safe state
/figma # Main Figma integration menu
/figma-auth # Authenticate with Figma
/figma-list # List available Figma files
/figma-select # Select Figma file
/figma-inspect # Inspect Figma components
/figma-generate # Generate code from designs
/figma-sync # Sync changes with Figma
/figma-create # Create new Figma components
/generate-agent <spec> # Create custom specialized agents
/setup-ollama # Install and configure Ollama
/setup-lmstudio # Configure LM Studio
/local-models # List available local models
/configure-routing # Configure smart model routing
/setup-mcps # Configure 25 MCP servers
/monitor # Launch real-time dashboard
/visual-test # Run visual regression tests
/docs # Open documentation hub
/configure-agentwise # Configure settings
/upload <file> # Upload documents (PDF, Word, Figma)
/image # Visual context file browser
/deploy # Deploy project to production
/update-agentwise # Update to latest version
📚 View Complete Commands Documentation for detailed usage and examples
# Create a new project directly
/create "an e-commerce platform with Next.js and Stripe"
# Monitor progress in real-time
/monitor
# Add features to active project
/task "add user authentication with OAuth"
Agentwise includes a comprehensive web-based monitoring dashboard that provides real-time visibility into your agent orchestration:
/monitor # Opens dashboard (auto-installs global command)
/monitor install # Manually install global command
/monitor global # Alternative install command
/monitor status # Check installation status
/monitor help # Show monitor command help
Note: Running /monitor without subcommands now automatically installs the global command if not present.
Global Command: After installation, use agentwise-monitor from anywhere:
agentwise-monitor # Start monitor from any directory
agentwise-monitor status # Check system status
Or manually:
cd src/monitor
./start.sh # Starts both WebSocket server and web UI
The monitoring dashboard is designed for local development use only. It runs on localhost and should not be exposed to external networks without additional security measures.
Transform any project idea into a fully configured, production-ready application with a single command:
/create-project "e-commerce platform with payments"
This powerful wizard handles:
Generate comprehensive project requirements from simple descriptions:
/requirements "social media platform" # Generate requirements
/requirements-enhance # Enhance existing requirements
/requirements-visualize # Create HTML documentation
/requirements-to-tasks # Convert to development tasks
Features:
Zero-configuration database setup with automatic type generation:
/database-wizard # Interactive setup
/database-setup supabase # Quick setup
/database-connect # Connect to existing
Capabilities:
Complete GitHub repository setup and management:
/github-setup # Complete GitHub setup
/github-sync # Sync code with GitHub
/github-secrets # Manage repository secrets
Features:
Continuous protection for your codebase:
/enable-protection # Enable all protection
/protection-status # View protection status
/security-report # Generate security report
Protection includes:
Choose how you want to work:
# Full setup with everything
/create-project "app" --with-database --with-github --with-protection
# Minimal local-only setup
/create-project "app" --local-only
# Add features later as needed
/database-wizard # Add database anytime
/github-setup # Add GitHub anytime
/enable-protection # Add protection anytime
| Command | Description | Example |
|---|---|---|
/create <idea> | Start new project from scratch | /create "social media dashboard" |
/create-plan <idea> | Collaborative planning session | /create-plan "mobile app" |
/projects | List and select projects | /projects |
/task <feature> | Add feature to active project | /task "add dark mode" |
/task-[project] <feature> | Add feature to specific project | /task-dashboard "add charts" |
/task-plan <feature> | Plan feature collaboratively | /task-plan "payment system" |
| Command | Description | Example |
|---|---|---|
/init-import | Import existing project | /init-import |
/task-import | Copy and integrate project | /task-import |
/upload <file> | Upload documents or design files | /upload ./design.fig components |
/clone-website <url> | Clone and customize websites | /clone-website https://example.com similar |
| Command | Description | Example |
|---|---|---|
/generate-agent <type> | Create custom agent | /generate-agent "security-specialist" |
/monitor [subcommand] | Monitor dashboard & global install | /monitor install |
/docs | Open local documentation hub | /docs |
| Command | Description | Example |
|---|---|---|
/figma connect | Connect to Figma Dev Mode | /figma connect |
/figma generate [name] | Generate component from Figma | /figma generate Button |
/figma sync | Sync design tokens & components | /figma sync |
/figma tokens [path] | Export design variables | /figma tokens ./tokens.json |
/figma image [path] | Capture design screenshot | /figma image ./design.png |
/figma rules [dir] | Generate design system rules | /figma rules ./design-system |
/figma status | Check Figma connection status | /figma status |
| Command | Description | Example |
|---|---|---|
/setup-ollama | Setup Ollama for local models | /setup-ollama |
/setup-lmstudio | Setup LM Studio integration | /setup-lmstudio |
/local-models | List available local models | /local-models |
/configure-routing | Configure model routing | /configure-routing optimize |
| Command | Description | Example |
|---|---|---|
/image | Visual context with file browser | /image |
/docs | Open comprehensive documentation hub | /docs |
/security-review | Run security analysis | /security-review |
/deploy | Deploy to production | /deploy production |
/rollback | Rollback deployment | /rollback |
agentwise/
├── .claude/ # Claude Code integration
│ ├── agents/ # Agent definitions (frontend, backend, etc.)
│ └── commands/ # Custom command handlers
├── src/ # Core system architecture
│ ├── agents/ # Dynamic agent management
│ ├── ai/ # AI prompt enhancement & optimization
│ ├── analytics/ # Performance & usage analytics
│ ├── backup/ # Project backup & restore system
│ ├── cli/ # Command-line interface handlers
│ ├── commands/ # Command implementations
│ │ ├── GlobalMonitorInstaller.ts # Cross-platform global commands
│ │ ├── MonitorCommand.ts # Monitor dashboard controller
│ │ ├── ImageCommand.ts # Visual context processing
│ │ ├── ModelCommands.ts # Local model management
│ │ └── UploadHandler.ts # Document & design file processing
│ ├── context/ # Project context & persistence
│ ├── learning/ # Self-improving agent capabilities
│ ├── mcp/ # MCP server integration (25 verified servers)
│ ├── models/ # Smart model routing & local model support
│ ├── monitor/ # Real-time dashboard (Next.js app)
│ │ ├── server/ # WebSocket server for live updates
│ │ └── src/ # Dashboard UI components
│ ├── monitoring/ # Task completion & progress tracking
│ ├── optimization/ # Context management and task coordination
│ ├── orchestration/ # Multi-agent coordination
│ ├── orchestrator/ # Agent orchestration & management
│ ├── project-registry/ # Project synchronization system
│ ├── projects/ # Project management utilities
│ ├── spec-templates/ # Project specification templates
│ ├── utils/ # Helper utilities & shared code
│ └── validation/ # Code, style, and tech stack validation
├── config/ # Configuration files
├── docs/ # Comprehensive documentation
│ ├── examples/ # Usage examples & tutorials
│ ├── monitor-command.md # Monitor command documentation
│ ├── custom-agents.md # Custom agent creation & management
│ ├── ci-cd-integration.md # CI/CD pipeline integration guide
│ ├── architecture.md # System architecture details
│ ├── mcp-integration.md # MCP server integration guide
│ └── smart-model-routing.md # Model routing documentation
├── installers/ # Platform-specific installers
└── workspace/ # Project workspaces (git-ignored)
└── [project-name]/ # Individual project directories
└── agent-todos/ # Agent task management per project
Agentwise includes 8 specialized agents, each with unique capabilities and MCP integrations:
/generate-agent "specialization" commandEvery agent in Agentwise has self-improvement capabilities:
.agent-knowledge directoryEach agent continuously improves, becoming more efficient over time!
The Knowledge Graph system provides semantic understanding of your entire codebase:
Combined with context sharing, achieves 15-30% total token reduction!
Automatically validates all claims made by AI agents:
Every agent claim is automatically tested and validated!
graph TD
A[User Command] --> B[Agentwise Orchestrator]
B --> C[Smart Model Router]
C --> D[Claude/Ollama/LM Studio]
B --> E[Dynamic Task Distributor]
E --> F[Tech Stack Validator]
F --> G[Agent Selection]
G --> H1[Frontend Specialist]
G --> H2[Backend Specialist]
G --> H3[Database Specialist]
G --> H4[DevOps Specialist]
G --> H5[Testing Specialist]
G --> H6[Designer Specialist]
G --> H7[Research Agent]
G --> H8[Custom Agents]
H1 --> I[MCP Integration]
H2 --> I
H3 --> I
H4 --> I
H5 --> I
H6 --> I
H7 --> I
H8 --> I
I --> J[25 Verified MCP Servers]
J --> K[Phase Controller]
K --> L[Self-Improving Agents]
L --> M[Performance Analytics]
M --> N[Token Optimizer 30-40%]
N --> O[Project Output]
Agentwise features comprehensive MCP (Model Context Protocol) integration with actually existing, verified servers:
# Frontend agent uses Figma MCP
/figma generate Button # Extracts button from Figma
# Database agent uses PostgreSQL MCP
# Automatically optimizes queries and suggests indexes
# Testing agent uses Playwright MCP
# Generates and runs E2E tests automatically
We love contributions! See CONTRIBUTING.md for guidelines.
git checkout -b feature/AmazingFeature)git commit -m 'Add some AmazingFeature')git push origin feature/AmazingFeature)# Clone your fork
git clone https://github.com/YOUR_USERNAME/agentwise.git
cd agentwise
# Install dependencies
npm install
# Run tests
npm test
# Build project
npm run build
✅ Comprehensive security audit completed in response to user concerns:
AGENTWISE_ANALYTICS=false to opt-out)Found a security issue? Please email vibephilcoding@gmail.com or create a private security advisory on GitHub.
10xFaster Development |
30-40%Token Reduction |
8+Parallel Agents |
25Verified MCPs |
Note: Agentwise was created on August 20, 2025
We're keeping Agentwise focused and lightweight. See ROADMAP.md for details.
Agentwise is developed and maintained through community support. Your donations help continue development and add new features!
Every donation, no matter the size, is greatly appreciated and helps keep Agentwise free for personal use! 🙏
We're planning to extend Agentwise to support additional AI CLI tools:
We take security seriously and have conducted comprehensive security audits with the following improvements:
If you discover a security vulnerability, please:
For more details, see our Security Policy.
Agentwise uses a dual license model to balance open-source freedom with sustainable development.
✅ Use Agentwise for FREE for:
❌ Only if you are:
FREE (Apache 2.0):
REQUIRES LICENSE:
Simple Rule: Using Agentwise as a TOOL = Free. Reselling Agentwise itself = License required.
See the LICENSE file for complete details and FAQ.
Built with ❤️ by Philip Ritmeester
1000+ skills curated from Anthropic, Vercel, Stripe, and other engineering teams
@user_bdd01887CommunityDesign enforcement with memory — keeps your UI consistent across a project
Universal SEO skill for Claude Code. 25 sub-skills + 18 sub-agents covering technical SEO, E-E-A-T, schema, GEO/AEO, bac
Route Claude Code traffic to any of 17 provider backends including free or local models
