Welcome to Agentwise documentation. Learn how to harness the power of multi-agent orchestration to accelerate your development workflow with verified 64.6% token reduction (up to 94.8% with 10 agents).
Get Agentwise up and running in minutes
Get startedYour first project with Agentwise
Get startedComplete guide to all Agentwise commands
Get startedCreate your own specialized AI agents
Get startedBefore getting started with Agentwise, ensure you have the following:
Agentwise now includes sandboxed execution with automatic permission handling:
# Works without any flags!
claude /create "my amazing app"
# Old way (still supported but not needed):
claude --dangerously-skip-permissions
✅ Automatic Permission Handling: Responds to prompts intelligently
✅ Workspace Sandboxing: Restricts execution to safe directories
✅ Smart Safety Modes: Configurable levels of automation
Copy this prompt and paste into Claude Code for automatic setup:
Please set up Agentwise by:
1. Clone: git clone https://github.com/VibeCodingWithPhil/agentwise.git ~/agentwise
2. Install: cd ~/agentwise && npm install && npm run build
3. Configure: Create workspace directory and set up global monitor
4. Start: node dist/index.js
5. Enable sandboxed execution (no --dangerously-skip-permissions needed)
Full setup: https://github.com/VibeCodingWithPhil/agentwise/blob/main/CLAUDE_CODE_SETUP_PROMPT.md
curl -fsSL https://raw.githubusercontent.com/VibeCodingWithPhil/agentwise/main/installers/install.sh | bash
Invoke-WebRequest -Uri "https://raw.githubusercontent.com/VibeCodingWithPhil/agentwise/main/installers/install.ps1" -OutFile "install.ps1"
Set-ExecutionPolicy -ExecutionPolicy RemoteSigned -Scope CurrentUser
.\install.ps1
Start Claude Code (no flags needed!)
claude
Agentwise handles permissions automatically
Verify installation
/help
You should see Agentwise commands listed.
Create your first project
/create "a todo app with React and Firebase"
Monitor progress
/monitor
Opens the dashboard at http://localhost:3001
Understanding how Agentwise orchestrates multiple AI agents for optimal performance:
Learn about token optimization and benchmarks
View metricsCreate specialized agents for your workflow
Create agentsExplore real-world usage examples
See examplesNeed help or want to contribute? Join our community: